Disparities in Access to Well-Qualified,Well-Supported Special Educators Across Higher- Versus Lower-Poverty Schools Over Time

Abstract

Scholars have documented long-standing disparities in access to well-qualified, well-supported teachers, including disparities in access to special education teachers (SETs), based on student socioeconomic status. In response, policy initiatives have aimed to incentivize teaching in higher-poverty schools. Thus, we examined changes over time in disparities between SETs’ demands and resources (including internal resources, such as qualifications, and school-based resources, such as adequate materials), using multiple waves of the nationally representative Schools and Staffing Survey. We found that, by one metric, disparities in certification have closed since 2000. However, SETs in higher poverty schools are significantly more likely to work in self-contained settings than those in lower-poverty schools, and disparities in school-based resources continue to be significant, such that SETs in higher-poverty schools were significantly more likely to teach in self-contained classes, rated teacher cooperation significantly lower, and reported having significantly weaker access to material resources.

Keywords

special educators school personnel equity working conditions

Strong teachers are crucial to students’ long-term outcomes (Chetty et al., 2014), yet scholars have documented long-standing disparities in access to well-qualified, well-supported teachers, based on the average socioeconomic status of students in a school (Simon & Johnson, 2015). These disparities represent a substantial inequity in the educational opportunities schools provide to students who live in higher- versus lower-poverty areas.

As with teachers in general, scholars have documented substantial gaps in access to well-qualified, well-supported special education teachers (SETs) based on the proportion of students in their school who qualify for free-and-reduced-price-lunches (FRPL; Mason-Williams, 2015). SETs are entrusted with crucial responsibilities to meet the needs of students with substantial learning and behavior needs, including providing intensive intervention to address foundational skill deficits and ensuring students continue to access and make progress in the general education curriculum (Sayeski et al., 2019). SETs’ capacity to fulfill this charge depends on the extent to which they have reasonable responsibilities (e.g., appropriate caseloads) and resources to fulfill those responsibilities (including internal resources, such as experience, and school-based resources, such as social support and materials; Billingsley et al., 2020).

Several studies have documented disparities. Fall and Billingsley (2008) analyzed the Study of Personnel Needs in Special Education (SPeNSE), a nationally representative survey of K12 SETs from 1999–2000, and found that SETs in higher-poverty districts (i.e., serving > 39% students living in poverty) were less likely to be fully certified, more likely to hold an emergency certificate, less likely to have completed student teaching, and less likely have had ≥ 10 weeks of student teaching. Similarly, analyzing the nationally representative Schools and Staffing Survey (SASS), from the 2003–2004 school year, Mason-Williams (2015) found that, in K12 schools in the highest quartile for the proportion of students eligible for FRPL, SETs were less likely to have extensive teacher preparation, less likely to hold degrees in special education, less likely to be certified in special education or any content area, and less likely to have > 3 years’ experience.

These studies indicate SETs in higher-poverty schools hold significantly weaker qualifications than SETs in lower-poverty schools, likely resulting in weaker internal resources to fulfill responsibilities (Mason-Williams et al., 2020). Thus, reasonable responsibilities and strong school-based resources (e.g., social support, materials) may be especially important for facilitating their capacity to fulfill their responsibilities (Billingsley et al., 2020). Yet, extant studies indicate these SETs are charged with significantly higher responsibilities (e.g., larger caseloads), and they have significantly fewer resources (e.g., administrative and collegial support) than their counterparts in lower-poverty schools (Fall & Billingsley, 2011).

Closing these gaps is an important policy priority, and initiatives at both state and federal levels have aimed to incentivize teachers to teach in higher-poverty schools (Rothstein, 2015). For example, the Every Student Succeeds Act (ESSA, 2015) requires states to develop Equity Plans, detailing how they will reduce disparities in access to well-qualified teachers. Initiatives to reduce disparities vary in nature, scope, and timing, but include (a) financial incentives to teach special education in high-poverty schools (e.g., Feng & Sass, 2017); (b) Teacher Education Assistance for College and Higher Education (TEACH) grants, which provide federal loans to pursue licensure, in exchange for a commitment to teach in high-poverty schools (Peyton et al., 2020); (c) evaluating teacher preparation programs based, in part, on the proportion of graduates in high-poverty schools (Tatto et al., 2016); (d) increasing support for teachers in high-poverty schools, particularly early in their careers (Simon & Johnson, 2015); and (e) supporting programs, such as Teach For America, that place teachers in high-poverty schools (Brewer et al., 2016). The extent to which these initiatives have, collectively, moved the needle on gaps in access to well-qualified, well-supported SETs is unknown.

Thus, the purpose of this investigation is to trace how gaps between SETs’ qualifications and supports in higher- versus lower-poverty schools have changed over time, using nationally representative data. This has important implications for ensuring equitable access to well-qualified, well-supported SETs. We do not aim to identify how specific changes in policy have contributed to closing gaps, but, rather, how these gaps have changed over time, in a context in which an array of policy initiatives have been enacted. This information is crucial for understanding if new policies are needed to address disparities.

Conceptual Framework: Conservation of Resources Theory

We use conservation of resources (COR) theory (Alarcon, 2011) as a conceptual foundation for examining changes in SETs’ qualifications and supports over time. Developed by organizational studies scholars, COR posits employees meet job demands (i.e., expectations or responsibilities of their work) by strategically drawing on resources available to them (Alarcon, 2011; Halbesleben et al., 2014). Resources include anything employees believe helps them meet the demands of their job, including internal resources (e.g., knowledge, skills gained through preparation), social resources (e.g., leaders’ support), and logistical resources (e.g., materials; Halbesleben et al., 2014). When demands are high, they may experience negative affective and job outcomes (e.g., burnout, quitting, poor performance) without appropriate resources (Alarcon, 2011). Meta-analyses affirm core tenets of COR theory, indicating it provides a useful schema for conceptualizing key supports for employees across professions (Alarcon, 2011; Halbesleben et al., 2014). Recent studies have used COR theory to conceptualize influences on SETs’ experiences and outcomes, and have obtained results consistent with the theory, indicating it is applicable to SETs (e.g., Bettini, Gilmour et al., 2020). Through the lens of COR theory, SETs’ efforts to effectively serve students with disabilities depend on reasonable job demands and adequate resources to meet demands, including internal resources (e.g., certification) and school-based resources (e.g., materials; Bettini, Gilmour et al., 2020).

Demands

Demands are the roles and responsibilities SETs are charged with fulfilling for their job, including SETs’ service delivery model, instructional responsibilities, caseload management, extra responsibilities, and addressing student challenges (Bettini, Gilmour et al., 2020).

Service Delivery Model

Service delivery model refers to the ways special education services are organized for the students whom a SET serves, and is often defined based on students’ placement on the continuum of least restrictive environments, including (a) co-teaching or consulting in general education settings; (b) resource teaching, in which SETs provide services by either pushing into general education, or pulling students to separate settings for intervention; and (c) teaching self-contained classes, providing most or all instruction in a separate specially-designed setting in a neighborhood school (Rozalski et al., 2010). Although some demands (e.g., paperwork) are common across service delivery models, extant research suggests demands on SETs may vary substantially based on the model in which they work (Embich, 2001). For example, research with elementary SETs indicates those in inclusive settings (e.g., co-teaching) may have more intensive collaborative responsibilities than those in separate settings (e.g., resource, self-contained; Bettini et al., 2021 Klingner & Vaughn, 2002), whereas those in self-contained settings may be relatively isolated (O’Brien et al., 2019).

Extant research indicates co-teaching has been increasing since 2000 (Gilmour et al., 2021), despite limited evidence that it is effective at promoting stronger student outcomes (Jones & Winters, 2020) and despite consistent evidence that SETs in co-taught settings often fulfill marginal instructional roles (Scruggs et al., 2007; Wexler et al., 2018). Simultaneously, self-contained service delivery models have been declining (OSEP, 2019). However, extant research provides limited insight into how service delivery models may vary based on school poverty. Examining administrative data from Massachusetts, Grindal et al. (2019) found high-poverty districts placed a higher proportion of students in restrictive settings than low-poverty districts. Analyzing a nationally representative sample of adolescents with mental health disorders, Green et al. (2020) found adolescents whose parents had lower educational attainment were more likely to receive services in a separate class or school. Neither study examined the service delivery model in which SETs worked, but they both indicate that SETs in schools serving more socioculturally marginalized students may be more likely to work in restrictive settings, while SETs serving more socioculturally privileged students may be more likely to work as co-teachers.

Instructional Responsibilities

Instruction is a core component of SETs’ work, and planning for and providing instruction constitute demands on SETs’ time and energy (Vannest & Hagan-Burke, 2010). SETs are often responsible for teaching students across multiple grades and subjects (O’Brien et al., 2019). Although SETs report feeling more positive emotions during instruction than during other aspects of their work (Jones & Youngs, 2012), and view instruction as core to their roles (Bettini et al., 2019), they also report feeling more overwhelmed when they are responsible for preparing more lessons or when they teach more subjects to a wider range of different grade levels (Bettini, Cumming et al., 2020). However, extant research provides no insights into whether SETs in higher-poverty settings may have greater or lesser instructional demands.

Caseload Management Responsibilities

Caseload management is a core responsibility, encompassing assessing student progress, completing paperwork for Individualized Education Programs (IEPs), and coordinating service delivery across service providers, which collectively take up substantial portions of K12 SETs’ time (Vannest & Hagan-Burke, 2010). Moreover, SETs consistently report that these demands are among the least satisfying aspects of their work (Jones & Youngs, 2012), with consistent associations with their intent or decisions to leave their jobs (e.g., Billingsley & Bettini, 2019).

Extant research provides little insight into whether caseload management responsibilities vary across higher- versus lower-poverty schools. Analyzing nationally representative data from 1999–2000, Fall and Billingsley (2011) found K12 SETs in higher-poverty districts served an average of 22 students; those in low-poverty districts served an average of 18 students. More recent figures are not available. Dewey et al. (2017) found SETs’ caseloads likely increased nationally through 2012, as the number of SETs employed in the United States declined while the number of students with disabilities increased, but they did not examine disparities across schools. No extant studies have, to our knowledge, examined if the disparities Fall and Billingsley (2011) identified have changed.

Extra Duties

Extra duties include tasks such as supervising students during non-instructional times (e.g., bus duty), which SETs report interfere with their teaching (Vannest & Hagan-Burke, 2010). Extant research indicates SETs who report more responsibilities that interfere with instruction are more likely to intend to leave (e.g., Bettini, Cumming et al., 2020). In their analysis of a nationally representative survey from 1999–2000, Fall and Billingsley (2011) found that K12 SETs in higher-poverty schools were significantly more likely to report that “routine duties and paperwork” interfered with teaching, but this item conflates extra duties with caseload management demands (i.e., paperwork), and no studies have confirmed these findings using more recent data.

Addressing Student Behavior Problems

The students who SETs are responsible for serving can be the most rewarding aspect of the job (Prather-Jones, 2011), but responsibility for addressing more significant student behavior challenges seems to present a greater demand on teachers (Feng, 2009), including SETs. Studies consistently indicate that SETs who serve more students with more significant behavior challenges are more likely to leave or intend to leave (e.g., Gilmour & Wehby, 2020). For example, in an analysis of the 2011–2012 SASS, Bettini, Gilmour et al. (2020) found K12 teachers’ ratings of “student problems” were significantly associated with intent to leave. The magnitude of the association was smaller for SETs than for general educators, but still significant and meaningful. Due to leadership instability (Béteille et al., 2012) and other resource constraints (Baker et al., 2020), higher-poverty schools may have more difficulty initiating and sustaining school-reform initiatives designed to systematically support student behavior, such as Positive Behavioral Interventions and Supports (PBIS; Molloy et al., 2013), potentially resulting in teachers experiencing more demands to address student behaviors on their own.

Resources

To fulfill their commitments to students, SETs depend on both internal resources and school-based resources, such as administrative support and cooperation among teachers.

Internal Resources: Qualifications

Qualifications are formal indicators a SET has had experiences (e.g., opportunities to learn) likely to increase their knowledge and skill to effectively teach students with disabilities. Often termed teacher characteristics (e.g., Feng & Sass, 2013), qualifications include certification and experience. Certification indicates a teacher has completed the requirements (typically coursework, licensure exams, and supervised teaching experience) deemed by their state as necessary to teach in a specific setting (Goldrick et al., 2014). Some evidence shows that the training indicated by special education certification does translate into more effective teachers (e.g., Feng & Sass, 2013), although not for all disability categories (Gilmour, 2020; Goldman & Gilmour, 2020). For example, Feng and Sass (2013) found that students with disabilities in Florida had slightly higher English language arts and math scores when taught by a special education certified versus a general education certified teacher. Gilmour and Wehby (2020) found that the association between the percentage of students with disabilities in teachers’ classes and their odds of attrition was completely attenuated when teachers were certified in special education. As previously noted, past research has identified disparities in access to certified special educators across higher- and lower-poverty schools (Mason-Williams, 2015).

General education research consistently finds that more experienced teachers are more effective at improving students’ academic outcomes (Feng & Sass, 2013; Kraft & Papay, 2014; Rockoff, 2004) and are at much lower risk of attrition than early career teachers (Nguyen et al., 2020). Research regarding the association between teaching experience and the outcomes of students with disabilities is more mixed. Feng and Sass (2013) found that students with disabilities in Florida had, on average, higher math and reading scores when they were taught by more experienced general or special education teachers, suggesting that as teachers gain more experience, they also gain knowledge of how to work with students who present different needs. In contrast, Theobald et al. (2020) found that reading scores among high school students with disabilities in Washington were not associated with teacher experience. Despite these divergent findings regarding the importance of teacher experience, Mason-Williams' (2015) findings, that students in less-resourced schools tend to have less experienced teachers, are worrisome.

School-Based Resources

SETs also depend on school-based resources, including social supports and material resources, to effectively meet demands (Bettini, Gilmour et al., 2020). Social supports from colleagues and administrators are especially essential for SETs because SETs’ work depends on the coordination of effort among multiple educators to ensure students’ services are consistent and well-coordinated across environments in the school (Brownell et al., 2010). SETs spend substantial proportions of their time collaborating with other educators (e.g., Vannest & Hagan-Burke, 2010). They depend on cooperation with colleagues to ensure students receive needed accommodations and modifications to the general education curriculum, to ensure they are able to access their students for intervention instruction (Bettini et al., 2021), and, when co-teaching, to ensure they can play a meaningful role in planning and providing instruction (Scruggs et al., 2007). SETs report experiencing stronger coordination when their administrators support them by establishing clear expectations for how all staff should be serving students with disabilities (e.g., Youngs et al., 2011); further, they also depend on administrators to communicate trust, provide feedback on their instruction, provide needed resources, and work with them to address student behavior and other challenges (Billingsley et al., 2017). Both collegial support and administrative support are consistently associated with SETs’ intent to stay in their schools and in the profession (Billingsley & Bettini, 2019), and research with general educators indicates that teachers who experience stronger collegial and administrative support may demonstrate stronger student achievement gains (Ronfeldt et al., 2015). No recent studies have compared SETs’ access to administrative support and teacher cooperation across higher- versus lower-poverty schools, but Fall and Billingsley (2011) found that SETs in high-poverty districts reported significantly weaker principal and collegial support than those in low-poverty districts.

To fulfill instructional responsibilities, SETs also depend on material resources, such as curricula, technology, and basic supplies (e.g., paper, pencils; Billingsley et al., 2020). Materials provide teachers practical tools needed to fulfill their responsibilities. For example, curricula support teachers in defining their instructional scope and sequence, as well as methods for teaching and assessing particular content (Siuty et al., 2018). Extant research on how curricula affects SETs’ instruction is limited (Bettini et al., 2016), but a growing body of research indicates teachers become more effective when they have strong curricular resources (Jackson & Makarin, 2016). Further, SETs are more likely to report that they can manage their workloads and that they intend to stay in their jobs when they report having strong material resources (e.g., Bettini, Gilmour et al., 2020). No recent studies have compared SETs’ access to material resources across higher- versus lower-poverty schools, but Fall and Billingsley found SETs in high-poverty districts were significantly less likely than teachers in low-poverty districts to report that “necessary materials are available when you need them” (p. 71).

Methods

To examine how disparities in SETs’ demands and resources across higher- versus lower-poverty schools have changed over time, we used data from the nationally representative Schools and Staffing Survey (SASS) and the newest iteration of SASS, the National Teacher Principal Survey (NTPS), both of which were collected and administered by the National Center for Education Statistics (NCES). Based on prior research (e.g., Fall & Billingsley, 2011) and the ways schools replicate broader inequities (e.g., Baker et al., 2020), we hypothesized that SETs in higher-poverty schools would experience higher demands and weaker resources than SETs in lower-poverty schools. However, we proposed no hypotheses regarding how these gaps may have changed over time, as no prior research has examined this, and the potential effects of policy changes are unknown; our analysis of changes over time was exploratory.

Data and Measures

The SASS and NTPS surveys consisted of nationally representative samples of schools, principals, and teachers in the United States, with more than 30,000 public school teachers in each wave. These surveys included a rich set of relevant teacher and school characteristics. For this study, we used the four most recent waves of SASS, 1999–2000, 2003–2004, 2007–2008, 2011–2012, and NTPS 2015–2016 data, and we focused entirely on traditional public school teachers (i.e., excluding teachers in private, charter, and alternative schools). We employed appropriate sampling weights to ensure results were representative at the national level. NCES is currently adjusting sampling weights for NTPS 2015. However, we had NCES approval to use the currently available weights for this study. Notably, our results are substantively similar without the use of the sampling weights; these results are available upon request. The overall sample size for the descriptive analysis was 19,690 unique SET observations representing 1,713,500 SET teachers. We considered teachers SETs when they reported special education as their main teaching assignment. Approximately 42% of the sample taught in elementary schools, 44% in secondary schools, and 14% taught in combined elementary and secondary schools. The numbers of SETs were fairly comparable from wave to wave, particularly for the SASS; however, for NTPS, the number of SETs in the sample dropped to 2,720 compared to ≅ 4,000 SETs in the SASS. We dropped less than 0.5% of SETs due to missing data.

Measures

We included variables related to teacher demands and both internal and school-based resources. For detailed descriptions of each variable, and the NTPS items, see Appendix Table 1.

School Poverty

Consistent with prior research (Nguyen & Redding, 2018), we classified higher- and lower-poverty schools based on the proportion of students eligible for FRPL, designating schools with ≥ 50% students eligible for FRPL as higher-poverty schools, and schools with < 50% students eligible for FRPL as lower-poverty schools. This was our independent variable. Of note, FRPL does not address other aspects of socioeconomic status, such as parent education and occupational prestige (Kincaid & Sullivan, 2017), but it was the only measure of socioeconomic status in the SASS datasets. Because it is widely used in research, FRPL has the advantage of permitting comparison with other studies.

Demands

To evaluate dependent variables related to SETs’ demands, we examined service delivery models, the number of grades SETs taught, their caseloads, the extent to which they experienced student behavior issues, and, as a measure of overall demands, the number of hours SETs reported working. SETs reported their service delivery model as teaching in a self-contained setting, co-teaching with another teacher, or providing pull-out or push-in services (which we referred to as a resource service delivery model). Self-contained SETs instructed the same group of students all or most of the day in multiple subjects, while co-teachers were jointly responsible for teaching the same group of students all or most of the day. SETs who provided pull-out or push-in instruct a small number of students released from their regular classes to address specific needs. Each type of service delivery model was indicated by a binary variable.

We examined additional demands as dependent variables. To address whether other duties (e.g., paperwork) interfered with their teaching, we used an item that asked SETs to rate, on a 4-point Likert scale, the extent to which they agreed that routine duties and paperwork interfered with their teaching, with higher ratings representing greater agreement. We captured the number of grades SETs taught by calculating the number of different grades they reported teaching. To capture SETs’ caseloads, we included the number of students with IEPs they reported teaching. Using factor analysis, we also evaluated student behavioral issues in the school based on SETs’ responses to six items that asked to what extent student tardiness, absenteeism, class cutting, and misbehavior were problems at their school, and whether students had threatened to injure them. Our results indicate moderate to good internal consistency, with Cronbach alphas ranging from 0.753 to 0.774 for each wave. We standardized this variable. Total hours SETs reported working was a continuous variable for which SETs reported hours spent on teaching and school-related activities during a typical full week, during, before, and after school, and on weekends.

Resources

We examined three dependent variables related to SETs’ internal resources: qualifications, teaching experience, and field of study. Certification was a binary variable indicating they did not have any certification such as standard, probationary, provisional or temporary, or emergency certificate, and 0 if they were certified. Experience was a continuous variable of the number of years’ experience the SET reported. Last, field of study was a binary variable where 1 indicated the SET had focused on special education as a major field of study in undergraduate or graduate education and 0 if their field of study was not special education.

We also captured three dependent variables related to school-based resources, which have often been used in prior research (e.g., Bettini, Gilmour et al., 2020; Conley & You, 2017), addressing cooperation among teachers, administrator support, and whether SETs had access to adequate materials. SETs responded to each question on a scale of 1 (strongly agree) to 4 (strongly disagree). We reverse coded these such that higher values indicate that SETs were more likely to agree that they experienced more cooperation, administrative support, or materials.

Analysis

Using regression analysis, we estimated an ordinary-least square (OLS) model to examine if higher-poverty school status was associated with each outcome of interest and whether the association changed significantly over time. Our main model was:

$Y_{i j t} = β_{0} + β_{1} H i g h e r P o v e r t y_{i j t} + β_{2} H i g h e r P o v e r t y_{i j t} * γ_{t} + γ_{t} + λ_{k} + ε_{i j t}$
(1)

The goal of this analysis was to examine differential changes in SETs’ resources and demands across higher- and lower-poverty schools, after accounting for time-invariant state-level differences. Y represented an outcome of interest, such as SET field of study, for teacher i from school j in year t. HigherPoverty was a binary indicator for whether the SET worked in a higher-poverty school, $+ γ_{t}$ was a wave indicator, $λ_{k}$ represents state fixed effects to account for unobserved heterogeneity across states and $e_{i j t}$ was the random error term. We employed heteroskedastic-robust clustered standard errors at the state level to account for the nested nature of the data. Results and significance levels were unchanged when we used robust standard errors instead of clustered standard errors. We used appropriate weights provided by NCES for each wave for nationally representative results. To substantiate regression results, we also tested these differences descriptively, using the earliest (2000) and latest (2012, 2016) waves (Table 3).

Table 3.
Demands and Resources of Special Education Teachers (SETs) by Higher- and Lower-Free-and-Reduced-Price-Lunches (FRPL) School Status.

Wave: 2000 Wave: 2012 Wave: 2016

Low FRPL High FRPL Diff Low FRPL High FRPL Diff Low FRPL High FRPL Diff

Demands

Self-contained class 0.31 0.43 0.12 0.24 0.36 0.12 0.25 0.34 0.09

Co-teaching 0.10 0.06 −0.04 0.16 0.17 0.01 0.14 0.12 −0.02

Resource delivery model 0.41 0.37 −0.04 0.43 0.32 −0.1** 0.45 0.41 −0.03 +

Total hours in school 47.99 47.91 −0.09 49.65 50.18 0.52 51.23 50.73 −0.51

Duties interfere 2.04 2.04 0.00 2.06 2.03 −0.03 2.12 2.08 −0.03

Number of grades taught 3.12 3.23 0.11 3.42 3.32 −0.1 3.32 3.43 0.11 +

Case load size 16.45 17.01 0.57 20.70 19.24 −1.46 22.72 20.55 −2.17*

Student issue (std) −0.24 −0.02 0.2 −0.28 −0.03 0.26 −0.24 0.12 0.36

Internal Resources

No certification 0.05 0.1 0.05 0.00 0.00 0.00 0.02 0.02 0.00

Experience 13.15 12.7 −0.46 13.64 13.04 −0.6 13.42 12.73 −0.69 +

SPED field of study 0.75 0.68 −0.07** 0.7 0.63 −0.06 + 0.7 0.65 −0.05 +

School-Based Resources

Teacher coop. 2.07 1.98 -.09 2.20 2.12 -.08* 2.23 2.15 −0.08*

Admin. support 2.15 2.07 −0.08 2.33 2.25 −0.08* 2.36 2.31 −0.05

Adequate materials 1.95 1.76 −0.19 2.20 2.00 −0.20 2.12 2.00 −0.12**

Observations 4,100 4,240 2,720

Note. Nationally-representative weights are employed. Sample sizes weighted to the nearest 10 in accordance with NCES non-disclosure rule.

⁺ p < 0.10, * p < 0.05, p < 0.01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS) and National Teacher and Principal Survey (NTPS)

Sensitivity Analyses

To assess robustness of findings, we conducted additional analyses using different cut points to designate schools as higher or lower poverty. We specifically replaced the binary variable indicating if a school had ≥ 50% students eligible for FRPL with a binary variable indicating if the school was in the top or bottom quartile of students qualifying for FRPL.

In a second set of sensitivity analyses, we replaced the FRPL indicator with an indicator of whether > 50% students in the school came from minoritized racial or ethnic groups. Although socioeconomic status and race are highly correlated in the United States, we chose to examine these additional models addressing differences across schools serving higher and lower proportions of racially or ethnically minoritized students to ensure our focus on socioeconomic status was not masking inequities by race or ethnicity. While we focused our main analysis on differences between higher- and lower-poverty schools, prior work has also found substantial differences in teachers’ experiences and outcomes in schools serving higher proportions of students of color (e.g., Johnson et al., 2012). We also refitted these models examining differences across schools in the top and bottom quartile of the percentage of students from racially minoritized groups.

Results

We identified important disparities in SETs’ demands and resources across higher- and lower-poverty schools. In Tables 1 and 2 and Figures 1 and 2, we present results from regression analyses that assess differential changes in qualifications, demands, and resources over time. In Table 3, we present descriptive differences between the earliest and latest waves in our sample (results for 2004 and 2008 waves are comparable and, due to space limitations, are available upon request). We discuss regression and descriptive results by each category of outcomes.

Figure 1.
Changes in Demands Over Time by Higher- and Lower-Free-and-Reduced-Price-Lunches (FRPL) School Status.

Figure 2.
Changes in Internal- and School-Based Resources Over Time by Higher- and Lower-Free-and-Reduced-Price-Lunches (FRPL) School Status.

Table 1.
Demands Over Time by Higher- and Lower-Free-and-Reduced-Price-Lunches (FRPL) School Status.

Self-Contained Class Co-Teaching Resource Delivery Model Total Hours Working Other Duties Interfere Number of Grades Taught Case Load Size Student Issues (std)

Higher FRPL 0.105 −0.038 −0.015 −0.227 0.010 0.092 0.156 0.224

(0.031) (0.012) (0.024) (0.756) (0.051) (0.102) (1.022) (0.055)

SASS 2004 −0.055 −0.023 0.148 1.501* 0.022 0.246 7.150 −0.090*

(0.014) (0.015) (0.022) (0.636) (0.051) (0.074) (0.766) (0.035)

SASS 2008 −0.032 0.009 0.047* 1.329* 0.022 0.255 −5.602 −0.118*

(0.019) (0.018) (0.022) (0.603) (0.023) (0.071) (0.772) (0.047)

SASS 2012 −0.064** 0.055* 0.022 1.808 0.030 0.336 4.359** −0.061

(0.022) (0.021) (0.028) (0.398) (0.045) (0.100) (0.714) (0.045)

SASS 2016 −0.045* 0.043 0.018 3.336 0.071⁺ 0.149* 6.232** −0.008

(0.022) (0.012) (0.022) (0.415) (0.042) (0.069) (0.859) (0.041)

Higher FRPL* −0.005 0.027 −0.034 −1.018 −0.034 −0.167 −3.201⁺ 0.009

SASS 2004 (0.029) (0.016) (0.034) (0.904) (0.088) (0.128) (1.708) (0.073)

Higher FRPL* −0.054 0.084⁺ −0.056* 0.783 −0.001 −0.201 −1.092 0.043

SASS 2008 (0.043) (0.044) (0.027) (0.735) (0.058) (0.124) (1.072) (0.075)

Higher FRPL* 0.002 0.047⁺ −0.070* 0.339 −0.032 −0.201 −2.051 0.059

SASS 2012 (0.029) (0.024) (0.033) (0.890) (0.061) (0.149) (1.722) (0.088)

Higher FRPL* −0.022 0.018 −0.003 −0.477 −0.042 −0.002 −2.578* 0.151*

SASS 2016 (0.035) (0.017) (0.031) (0.938) (0.058) (0.107) (1.280) (0.072)

Constant 0.306 0.102 0.404 48.019 2.035 3.120 16.586 −0.233

(0.016) (0.011) (0.016) (0.363) (0.028) (0.058) (0.387) (0.026)

Observations 19,690 19,690 19,690 19,690 19,690 19,690 19,690 19,690

Note. Nationally-representative weights are employed. Heteroskedastic-robust standard errors are in parentheses. All models include state fixed effects.

⁺ p < .10, * p < .05, p < .01

Adapted from the United States Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS) and National Teacher and Principal Survey (NTPS)

Demands

We did not identify significant differences regarding the number of hours worked, whether other duties interfered with SETs’ work, number of grades taught, or case load size (except in 2016; Table 1). However, there were two significant and substantive differences worth noting. First, SETs in higher-poverty schools were significantly more likely to teach in self-contained classes than SETs in lower-poverty schools (Figure 1, top left). For example, in 2016, SETs in higher-poverty schools were 9 percentage points more likely to teach in self-contained classes. Other service delivery models did not significantly differ across settings (Table 1). Second, SETs in higher-poverty schools consistently reported more student behavior issues than those in lower-poverty schools (Figure 1, bottom middle); there was a 0.36 standard deviation (p < .01) difference in perceptions of student behavior issues between SETs in higher and lower-poverty schools in 2016 (Table 3). Group comparisons (Table 3) aligned with regression results.

Resources

Next, we examined differences in internal and school-based resources across settings. In 2000, a greater proportion of SETs were uncertified in higher-poverty schools than in lower-poverty schools (Figure 2, top left; Table 2). There was an overall decline in the proportion of uncertified teachers, with a greater decline in higher-poverty schools, resulting in a closing of the certification gap (Figure 2, top left). SETs in lower-poverty schools tended to have slightly more experience than SETS in higher-poverty schools, but these differences were not statistically significant and did not appear to change over time (Figure 2, top middle; Table 2). Special education as a field of study declined from 2000 to 2008 in both school settings, but began to rebound after 2008, although not to the percentages present in 2000 (Figure 2, top right; Table 2). Differences in special education as a field of study across school settings were consistently non-significant (Table 2). Of school-based resources, administrator support was significantly different overall (Table 2); SETs in higher-poverty schools reported having less administrator support than SETs in lower-poverty schools. SETs in higher-poverty schools reported less teacher cooperation than SETs in lower-poverty schools in 2012 and 2016 (Table 1). Across all waves, SETs in higher-poverty schools reported less access to materials (Figure 2). Descriptive group comparisons (Table 3) generally aligned with regression results.

Table 2.
Resources Over Time by Higher- and Lower-Free-and-Reduced-Price-Lunches (FRPL) School Status.

No Certification Experience SPED Field of Study Teacher Cooperation Administrative Support Adequate Materials

Higher FRPL 0.043 −0.393 −0.028 −0.104 + −0.130* −0.218

(0.015) (0.661) (0.021) (0.055) (0.057) (0.072)

SASS 2004 −0.040 0.404 −0.044* 0.178 0.242 0.164

(0.008) (0.377) (0.017) (0.032) (0.034) (0.030)

SASS 2008 −0.043 0.260 −0.112 0.136 0.281 0.268

(0.008) (0.364) (0.024) (0.035) (0.039) (0.057)

SASS 2012 −0.047** 0.443 −0.065* 0.134 0.187 0.258

(0.008) (0.326) (0.025) (0.030) (0.042) (0.050)

SASS 2016 −0.038 0.176 −0.049* 0.149 0.194 0.162**

(0.008) (0.454) (0.022) (0.030) (0.040) (0.052)

Higher FRPL SASS 2004 −0.034 0.251 0.009 0.114 0.003 0.030

(0.014) (0.660) (0.028) (0.074) (0.057) (0.064)

Higher FRPL SASS 2008 −0.039 0.676 0.027 0.078 0.011 −0.002

(0.016) (0.800) (0.036) (0.058) (0.070) (0.086)

Higher FRPL* SASS 2012 −0.046** 0.022 0.009 0.018 0.011 −0.005

(0.015) (0.900) (0.033) (0.069) (0.074) (0.102)

Higher FRPL SASS 2016 −0.043 −0.202 0.002 0.012 0.054 0.078

(0.016) (0.832) (0.029) (0.066) (0.063) (0.075)

Constant 0.053 13.136 0.739 2.074 2.163 1.955

(0.007) (0.294) (0.013) (0.021) (0.026) (0.037)

Observations 19690 19690 19690 19690 19690 19690

Note. Nationally-representative weights are employed. Heteroskedastic-robust standard errors are in parentheses. All models include state fixed effects.

⁺ p < .10, * p < .05, ** p < .01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS) and National Teacher and Principal Survey (NTPS)

Additional Analyses

We conducted supplementary analyses including (1) a sensitivity analysis using different cut-points to define higher-poverty versus lower-poverty schools; and (2) similar analyses, using the proportion of students from minoritized racial/ethnic backgrounds as the independent variable. First, we conducted a sensitivity analysis related to how we categorized higher- and lower-poverty schools. We compared schools in the highest and lowest quartiles of FRPL to ensure results were not sensitive to the cut-point used. Results were substantively similar to our main specification; in some instances, differences were even starker (Appendix Figures 1–2).

Next, we examined differences across schools serving higher and lower percentages of racially or ethnically minoritized students. Results were substantively similar to our main analysis (Appendix Tables 2–4). One notable difference was that SETs in schools serving smaller proportions of racially or ethnically minoritized students more often taught in a resource service delivery model than SETs in schools serving a larger proportions of White students. The remaining results were substantively similar to results from the analyses comparing higher- versus lower-poverty schools (Appendix Figures 3–4), with the exception of caseload sizes, which were more similar across higher and lower minority schools than higher- and lower-poverty schools. When we used the highest and lowest quartiles for schools serving students from racially or ethnically minoritized backgrounds, we found substantively similar results to our comparison between higher-poverty and lower-poverty schools, with two exceptions (Appendix Figures 5–6). SETs in the highest quartile were significantly less likely to teach in resource delivery model and taught fewer grade levels than SETs in the lowest quartile. In sum, across different operationalizations of schools’ sociocultural privilege/marginalization, we consistently found SETs had systematically different demands and resources, almost always to the detriment of SETs whose schools serve more students from socioculturally marginalized backgrounds.

Discussion

Disparities in access to strong teachers is a crucial source of inequity between students who attend higher- versus lower-poverty schools, and a number of policy initiatives have aimed to reduce these disparities (Mason-Williams et al., 2020). In this study, we sought to understand how disparities in SETs’ demands and resources have changed between 2000 and 2016, using nationally representative data. Our most encouraging finding was that disparities in SET certification closed; in 2000, SETs in higher-poverty schools were significantly less likely to be certified than SETs in lower-poverty schools, but they were equally likely to be certified in both settings in 2016. Note that, due to differences in wording of the special education certification item across SASS waves, results are not specific to special education certification, which has been shown to be especially important for SETs (Gilmour & Wehby, 2020); it is possible we might have obtained different results had we been able to examine special education certification. Nevertheless, this is an encouraging finding, indicating that, by one metric, socioeconomic gaps in access to certified SETs may be closing. We cannot draw conclusions about why this occurred, but policy changes over this time did specifically target increasing access to certified teachers (e.g., the No Child Left Behind Act’s highly qualified teacher requirements).

However, other disparities remained, as we found significant differences in both demands and school-based resources, almost uniformly privileging lower-poverty schools. Compared to SETs in lower-poverty schools, SETs in higher-poverty schools were significantly more likely to teach in self-contained classes, they rated student problems as significantly more challenging, they rated teacher cooperation significantly lower, and they reported having significantly weaker access to material resources. We considered the possibility that these variables could all be explained by an underlying factor, such as schools’ commitment to or progress in inclusive school reform. However, correlations among these variables were low (Appendix Table 5), suggesting that this possibility is unlikely and that these are distinct disparities. Further, results were robust to multiple alternative specifications of the model. These findings raise substantial concerns that SETs in higher-poverty schools may be tasked with meeting substantially more challenging demands, with fewer school-based resources than SETs in lower-poverty schools. COR theory suggests that these disparities likely have significant detrimental effects on SETs’ longevity in higher-poverty schools and on the services SETs may be able to provide to students with disabilities in higher-poverty schools.

Of note, disparities between higher- and lower-poverty schools were primarily related to school-based resources, not the resources teachers bring with them (e.g., certification). Policy to date has often focused on getting enough qualified teachers into higher-poverty schools, using, for example, State Equity Plans and financial incentives to ensure students in higher- and lower-poverty schools are equally likely to be taught by well-qualified teachers. Our results suggested that these initiatives may be incomplete if they do not also consider organizational conditions in schools, including inequities in demands that are then placed on these teachers, and resources that they can use to meet demands. If two students are equally likely to have a well-qualified teacher, but one teacher is being asked to do more work with less support, then those students may not have equitable access to the benefits of their teachers’ knowledge and skill. Placing more qualified teachers in higher-poverty schools is crucial to ensuring students have equitable access to strong educational opportunities, but, given long-standing evidence of the essential role of organizational conditions in fostering teachers’ success (e.g., Newmann et al., 2001; Ronfeldt et al., 2015), it is also likely insufficient; such efforts should also emphasize systematically reducing demands and improving resources in higher-poverty schools. Further, although school-based resources were generally improving over time, disparities between higher- and lower-poverty schools remained, indicating that closing disparities across school settings will require much higher investments in higher-poverty schools.

Of note, our findings regarding differences in SETs’ service delivery models parallel prior studies indicating that students in higher-poverty schools are less likely to be included in general education settings (e.g., Grindal et al., 2019). Understanding why these differences occur will be important for future research. Possibilities include (a) higher leader turnover (Béteille et al., 2012), which could limit higher-poverty schools’ capacity to initiate and sustain inclusive school reform (Sindelar et al., 2006); (b) expectations for parent engagement in special education that are misaligned with the resources and cultural expectations of many students attending higher-poverty schools (Kalyanpur et al., 2000) and (c) other resource disparities (e.g., Baker et al., 2020), which could lead schools to adopt service delivery models perceived as more efficient.

Limitations

We were constrained by variables consistently present in SASS and NTPS datasets; other variables (e.g., special education certification, planning time, professional development, coaches) would have been valuable to include but were either absent or phrased inconsistently across waves, precluding comparisons over time. Similarly, variables for service delivery model were not clearly defined, and different SETs may have interpreted these terms in different ways. Extant studies suggest self-contained service delivery models may be operationalized quite differently across schools (Bettini et al., 2021). More nuanced items would be useful for parsing this, but such items are not included in the datasets. Likewise, SASS and NTPS do not ask about student disability, nor about other key socioeconomic indicators, such as parent education.

Implications for Policy and Practice

ESSA (2015) currently requires states to develop Equity Plans to close gaps across multiple areas. Our results indicate states’ plans should be attentive to both inequities in SETs’ qualifications, and also inequities in the demands and resources SETs experience in schools. To identify these inequities, states need some form of systematic data collection to track teachers’ ratings of demands and resources across schools. Equity Plans offer an opportunity to implement these systems. Schools are required to measure and report “school quality and student success,” with options to include measures of student and educator engagement and school safety (Cook-Harvey et al., 2016, p. 7). Further, ESSA expands measures of students’ opportunities to learn, to include a focus on school resources including safe facilities, curricular materials, and teacher professional development (Cook-Harvey et al., 2016). Currently, it appears states’ plans focus primarily on students’ experiences of the school climate (Kostyo et al., 2018); state policymakers should acknowledge teachers’ experiences of their demands and resources as a component of school climate. Surveying teachers about their demands and resources fulfills these requirements, and would shed further light on disparities across school settings. Policymakers should thus consider adding teacher surveys to their Equity Plans, and using results of these surveys to collaborate with higher-poverty schools and districts to identify specific demands and resources that are particularly problematic for their SETs, and provide resources to build capacity around these demands and resources. For example, if a district’s data indicates student behavior challenges are particularly problematic, the state could provide support for initiating use of school-wide prevention systems (e.g., PBIS; Molloy et al., 2013). Similarly, if a school’s data indicate curricular resources are a particular challenge, the state could promote strong access to curricular resources and coherence in the district’s instructional program (Newmann et al., 2001) by helping district personnel select instructional curricula for students with disabilities that are both evidence-based and well-aligned with existing curricular norms and practices in the district.

Implications for Research

Future research should replicate and extend our results, examining whether they hold when using other datasets and analytic methods, and exploring other differences such as differences across elementary versus secondary schools. Prior work on reducing disparities in special education has often focused on access to qualified SETs (Mason-Williams et al., 2020). Our findings suggest other disparities warrant more attention, and should be the focus of future research. Reducing disparities will require that they first be systematically measured in ways that capture working conditions salient to SETs. Thus, scholars should consider conducting measurement research, validating tools states could use to systematically track SETs’ demands and resources across schools. Such tools should include measures of the constructs we examined, as well as other aspects of organizational conditions that prior research indicates are important, such as the match between SETs’ certification and their in-service role (Theobald et al., 2020), access to strong professional development (Kennedy, 2016), planning time (Bettini, Cumming et al., 2020), and mentors or coaches with relevant expertise (e.g., Cornelius et al., 2020).

Future research is also urgently needed to understand how to close these disparities. Scholars should consider using states’ Equity Plans (mandated by ESSA, 2015) to identify state policies aimed at improving supports for teachers in higher poverty schools, and testing their effects on disparities in teachers’ working conditions. Researchers could also partner with states and large districts, using research-practice partnership models to test potential strategies for reducing inequities in teachers’ support systems. Building stronger organizational conditions in higher-poverty schools will be necessary to close these gaps; thus, scholars should draw on research about how high poverty schools and districts foster strong organizational conditions (e.g., skilled leadership, coherent instructional programs; e.g., Bryk, 2010).

Such efforts would also likely be enhanced by more deeply understanding why these disparities exist. For example, how do other disparities (e.g., principals’ qualifications and retention [Béteille et al., 2012]; financial resources [Baker et al., 2020]) contribute? Research on the roots of disparities in working conditions could provide crucial insight into potential levers for providing SETs with more equitable demands and resources to serve their students.

Supplemental Material

sj-docx-1-ec-10.1177_00144029211024137 - Supplemental material for Disparities in Access to Well-Qualified, Well-Supported Special Educators Across Higher- Versus Lower-Poverty Schools Over Time

Supplemental material, sj-docx-1-ec-10.1177_00144029211024137 for Disparities in Access to Well-Qualified, Well-Supported Special Educators Across Higher- Versus Lower-Poverty Schools Over Time by Elizabeth Bettini, Tuan D. Nguyen, Allison F. Gilmour and Christopher Redding in Exceptional Children

Supplemental Material

sj-docx-2-ec-10.1177_00144029211024137 - Supplemental material for Disparities in Access to Well-Qualified, Well-Supported Special Educators Across Higher- Versus Lower-Poverty Schools Over Time

Supplemental material, sj-docx-2-ec-10.1177_00144029211024137 for Disparities in Access to Well-Qualified, Well-Supported Special Educators Across Higher- Versus Lower-Poverty Schools Over Time by Elizabeth Bettini, Tuan D. Nguyen, Allison F. Gilmour and Christopher Redding in Exceptional Children

	Wave: 2000	Wave: 2012	Wave: 2016
Demands
Self-contained class	0.31	0.43	0.12**	0.24	0.36	0.12**	0.25	0.34	0.09**
Co-teaching	0.10	0.06	−0.04**	0.16	0.17	0.01	0.14	0.12	−0.02
Resource delivery model	0.41	0.37	−0.04	0.43	0.32	−0.1**	0.45	0.41	−0.03 +
Total hours in school	47.99	47.91	−0.09	49.65	50.18	0.52	51.23	50.73	−0.51
Duties interfere	2.04	2.04	0.00	2.06	2.03	−0.03	2.12	2.08	−0.03
Number of grades taught	3.12	3.23	0.11	3.42	3.32	−0.1	3.32	3.43	0.11 +
Case load size	16.45	17.01	0.57	20.70	19.24	−1.46	22.72	20.55	−2.17*
Student issue (std)	−0.24	−0.02	0.2**	−0.28	−0.03	0.26**	−0.24	0.12	0.36**
Internal Resources
No certification	0.05	0.1	0.05**	0.00	0.00	0.00	0.02	0.02	0.00
Experience	13.15	12.7	−0.46	13.64	13.04	−0.6	13.42	12.73	−0.69 +
SPED field of study	0.75	0.68	−0.07**	0.7	0.63	−0.06 +	0.7	0.65	−0.05 +
School-Based Resources
Teacher coop.	2.07	1.98	-.09	2.20	2.12	-.08*	2.23	2.15	−0.08*
Admin. support	2.15	2.07	−0.08	2.33	2.25	−0.08*	2.36	2.31	−0.05
Adequate materials	1.95	1.76	−0.19**	2.20	2.00	−0.20**	2.12	2.00	−0.12**
Observations	4,100	4,240	2,720

	Self-Contained Class	Co-Teaching	Resource Delivery Model	Total Hours Working	Other Duties Interfere	Number of Grades Taught	Case Load Size	Student Issues (std)
Higher FRPL	0.105**	−0.038**	−0.015	−0.227	0.010	0.092	0.156	0.224**
	(0.031)	(0.012)	(0.024)	(0.756)	(0.051)	(0.102)	(1.022)	(0.055)
SASS 2004	−0.055**	−0.023	0.148**	1.501*	0.022	0.246**	7.150**	−0.090*
	(0.014)	(0.015)	(0.022)	(0.636)	(0.051)	(0.074)	(0.766)	(0.035)
SASS 2008	−0.032	0.009	0.047*	1.329*	0.022	0.255**	−5.602**	−0.118*
	(0.019)	(0.018)	(0.022)	(0.603)	(0.023)	(0.071)	(0.772)	(0.047)
SASS 2012	−0.064**	0.055*	0.022	1.808**	0.030	0.336**	4.359**	−0.061
	(0.022)	(0.021)	(0.028)	(0.398)	(0.045)	(0.100)	(0.714)	(0.045)
SASS 2016	−0.045*	0.043**	0.018	3.336**	0.071⁺	0.149*	6.232**	−0.008
	(0.022)	(0.012)	(0.022)	(0.415)	(0.042)	(0.069)	(0.859)	(0.041)
Higher FRPL*	−0.005	0.027	−0.034	−1.018	−0.034	−0.167	−3.201⁺	0.009
SASS 2004	(0.029)	(0.016)	(0.034)	(0.904)	(0.088)	(0.128)	(1.708)	(0.073)
Higher FRPL*	−0.054	0.084⁺	−0.056*	0.783	−0.001	−0.201	−1.092	0.043
SASS 2008	(0.043)	(0.044)	(0.027)	(0.735)	(0.058)	(0.124)	(1.072)	(0.075)
Higher FRPL*	0.002	0.047⁺	−0.070*	0.339	−0.032	−0.201	−2.051	0.059
SASS 2012	(0.029)	(0.024)	(0.033)	(0.890)	(0.061)	(0.149)	(1.722)	(0.088)
Higher FRPL*	−0.022	0.018	−0.003	−0.477	−0.042	−0.002	−2.578*	0.151*
SASS 2016	(0.035)	(0.017)	(0.031)	(0.938)	(0.058)	(0.107)	(1.280)	(0.072)
Constant	0.306**	0.102**	0.404**	48.019**	2.035**	3.120**	16.586**	−0.233**
	(0.016)	(0.011)	(0.016)	(0.363)	(0.028)	(0.058)	(0.387)	(0.026)
Observations	19,690	19,690	19,690	19,690	19,690	19,690	19,690	19,690

	No Certification	Experience	SPED Field of Study	Teacher Cooperation	Administrative Support	Adequate Materials
Higher FRPL	0.043**	−0.393	−0.028	−0.104 +	−0.130*	−0.218**
	(0.015)	(0.661)	(0.021)	(0.055)	(0.057)	(0.072)
SASS 2004	−0.040**	0.404	−0.044*	0.178**	0.242**	0.164**
	(0.008)	(0.377)	(0.017)	(0.032)	(0.034)	(0.030)
SASS 2008	−0.043**	0.260	−0.112**	0.136**	0.281**	0.268**
	(0.008)	(0.364)	(0.024)	(0.035)	(0.039)	(0.057)
SASS 2012	−0.047**	0.443	−0.065*	0.134**	0.187**	0.258**
	(0.008)	(0.326)	(0.025)	(0.030)	(0.042)	(0.050)
SASS 2016	−0.038**	0.176	−0.049*	0.149**	0.194**	0.162**
	(0.008)	(0.454)	(0.022)	(0.030)	(0.040)	(0.052)
Higher FRPL *SASS 2004	−0.034*	0.251	0.009	0.114	0.003	0.030
	(0.014)	(0.660)	(0.028)	(0.074)	(0.057)	(0.064)
Higher FRPL *SASS 2008	−0.039*	0.676	0.027	0.078	0.011	−0.002
	(0.016)	(0.800)	(0.036)	(0.058)	(0.070)	(0.086)
Higher FRPL* SASS 2012	−0.046**	0.022	0.009	0.018	0.011	−0.005
	(0.015)	(0.900)	(0.033)	(0.069)	(0.074)	(0.102)
Higher FRPL *SASS 2016	−0.043*	−0.202	0.002	0.012	0.054	0.078
	(0.016)	(0.832)	(0.029)	(0.066)	(0.063)	(0.075)
Constant	0.053**	13.136**	0.739**	2.074**	2.163**	1.955**
	(0.007)	(0.294)	(0.013)	(0.021)	(0.026)	(0.037)
Observations	19690	19690	19690	19690	19690	19690

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

The authors received no financial support for the research, authorship and/or publication of this article.

Supplemental Material

Supplemental material for this article is available online.

Manuscript received November 2020; accepted May 2021.

References

Alarcon

G. M

. (2011). A meta-analysis of burnout with job demands, resources, and attitudes. Journal of Vocational Behavior, 79(2), 549–562. https://doi.org/10.1016/j.jvb.2011.03.007

Baker

B. D.

Srikanth

Cotto

Jr. Green

P. C.

, III (2020). School funding disparities and the plight of latinx children. Education Policy Analysis Archives, 28(135), 1–26. https://doi.org/10.14507/epaa.28.5282

Bettini

Crockett

Brownell

M. T.

Merrill

. (2016). Relationships between working conditions and special educators’ instruction. Journal of Special Education, 50(3), 178–190. https://doi.org/10.1177/0022466916644425

Bettini

Cumming

M. M.

O’Brien

K. M.

Brunsting

N. C.

Ragunathan

Sutton

Chopra

. (2020). Predicting special educators’ intent to continue teaching students with emotional or behavioral disorders in self-contained settings. Exceptional Children, 86(2), 209–228. https://doi.org/10.1177/0014402919873556

Bettini

Gilmour

A. F.

William

T. O.

Billingsley

. (2020). Predicting special and general educators’ intent to continue teaching using conservation of resources theory. Exceptional Children, 86(3), 310–329. https://doi.org/10.1177/0014402919870464

Bettini

Morris-Mathews

Lillis

Meyer

Shaheen

Kaler

Brunsting

N. C.

. (2021). Special educators‘ roles in inclusive schools. Invited chapter for J. McLeskey, F. Spooner, B. Algozzine, & N.L. Waldron (Eds.), Handbook of Effective, Inclusive Elementary Schools: Research and Practice, Routledge, 43–76..

Bettini

Wang

Cumming

Kimerling

Schutz

. (2019). Special educators’ experiences of roles and responsibilities in self-contained classes for students with emotional/behavioral disorders. Remedial and Special Education, 40(3), 177–191. http://dx.https://doi.org/10.1177/0741932518762470

Béteille

Kalogrides

Loeb

. (2012). Stepping stones: Principal career paths and school outcomes. Social Science Research, 41(4), 904–919. https://doi.org/10.1016/j.ssresearch.2012.03.003

Billingsley

Bettini

. (2019). Special education teacher attrition and retention: A review of the literature. Review of Educational Research, 89(5), 697–744. https://doi.org/10.3102/0034654319862495

10.

Billingsley

Bettini

Morris-Mathews

H. M.

McLeskey

. (2020). Improving working conditions to support special educators’ effectiveness: A call for leadership. Teacher Education and Special Education, 43(1), 7–27. https://doi.org/10.1177/0888406419880353

11.

Billingsley

McLeskey

Crockett

J. B.

(2017). Principal leadership: Moving toward inclusive and high-achieving schools for students with disabilities (Document No. IC- 8). University of Florida, Collaboration for Effective Educator, Development, Accountability, and Reform Center. https://ceedar.education.ufl.edu/portfolio/principal-leadership-moving-towards-inclusive-and-high-achieving-schools-for-students-with-disabilities/

12.

Brewer

Kretchmar

Sondel

Ishmael

Manfra

. (2016). Teach for America's Preferential treatment: School district contracts, hiring decisions, and employment practices. Education Policy Analysis Archives, 24(15). https://doi.org/10.14507/epaa.24.1923

13.

Brownell

M. T.

Sindelar

P. T.

Kiely

M. T.

Danielson

L. C

. (2010). Special education teacher quality and preparation: Exposing foundations, constructing a new model. Exceptional Children, 76(3), 357–377. https://doi.org/10.1177/001440291007600307

14.

Bryk

. (2010). Organizing schools for improvement. Phi Delta Kappan, 91(7), 23–30. https://doi.org/10.1177/003172171009100705

15.

Chetty

Friedman

J. N.

Rockoff

J. E

. (2014). Measuring the impacts of teachers II: Teacher value-added and student outcomes in adulthood. American Economic Review, 104(9), 2633–2679. http://dx.https://doi.org/10.1257/aer.104.9.2633

16.

Conley

You

. (2017). Key influences on special education teachers’ intentions to leave: The effects of administrative support and teacher team efficacy in a mediational model. Educational Management Administration & Leadership, 45(3), 521–540. https://doi.org/10.1177/1741143215608859

17.

Cook-Harvey

C. M.

Darling-Hammond

Lam

Mercer

Roc

(2016). Equity and ESSA: Leveraging educational opportunity through the every student succeeds Act . Learning Policy Institute.

18.

Cornelius

K. E.

Rosenberg

Sandmel

K. N.

(2020). Examining the impact of professional development and coaching on mentoring of novice special educators. Action in Teacher Education, 42(3), 253–270. https://doi.org/10.1080/01626620.2019.1638847

19.

Dewey

Sindelar

P. T.

Bettini

Boe

E. E.

Rosenberg

M. S.

Leko

. (2017). Explaining the decline in special education teacher employment from 2005 to 2012. Exceptional Children, 83(3), 315–329. https://doi.org/10.1177/0014402916684620

20.

Embich

J. L

. (2001). The relationship of secondary special education teachers’ roles and factors that lead to professional burnout. Teacher Education and Special Education, 24(1), 58–69. https://doi.org/10.1177/088840640102400109

21.

Every Student Succeeds Act of 2015 , Pub. L. No. 114-95. (2015).

22.

Fall

A. M.

Billingsley

B. S.

(2008). Disparities in teacher quality among early career special educators in high-and low-poverty districts. Personnel Preparation, 181-208. Emerald Group Publishing Limited. https://doi.org/10.1016/S0735-004X(08)00007-4

23.

Fall

A. M.

Billingsley

B. S

. (2011). Disparities in work conditions among early career special educators in high-and low-poverty districts. Remedial and Special Education, 32(1), 64–78. https://doi.org/10.1177/0741932510361264

24.

Feng

. (2009). Opportunity wages, classroom characteristics, and teacher mobility. Southern Economic Journal, 75(4), 1165–1190. https://www-jstor-org-443.web.bisu.edu.cn/stable/27751438

25.

Feng

Sass

T. R

. (2013). What makes special-education teachers special? Teacher training and achievement of students with disabilities. Economics of Education Review, 36, 122–134. https://doi.org/10.1016/j.econedurev.2013.06.006

26.

Feng

Sass

T. R

. (2017). Teacher quality and teacher mobility. Education Finance and Policy, 12(3), 396–418. https://doi.org/10.1162/EDFP_a_00214

27.

Gilmour

A. F

. (2020). Teacher certification area and the academic outcomes of students with learning disabilities or emotional/behavioral disorders. The Journal of Special Education, 54(1), 40–50. https://doi.org/10.1177/0022466919849905

28.

Gilmour

A. F.

Wehby

J. H

. (2020). The association between teaching students with disabilities and teacher turnover. Journal of Educational Psychology, 112(5), 1042–1060. https://doi.org/10.1037/edu0000394

29.

Gilmour

Nguyen

Redding

Bettini

(2021). The shifting context of special education teachers’ work and its relationship with retention. Working Paper.

30.

Goldman

S. E.

Gilmour

A. F

. (2020). Educating students with autism Spectrum disorders: Is teacher certification area associated with academic outcomes? Journal of Autism and Developmental Disorders, 51, 1–14. https://doi.org/10.1007/s10803-020-04561-w

31.

Goldrick

Sindelar

Zabala

Hirsch

. (2014). The role of state policy in preparing educators to meet the learning needs of students with disabilities (Document No . PA-1). Collaboration for Effective Educator, Development, Accountability, and Reform Center.

32.

Green

J. G.

McLaughlin

K. A.

Alegria

Bettini

Gruber

Hoagwood

Le Tai

Sampson

Zaslavsky

A. M.

Xuan

Kessler

R. C.

(2020). Associations of sociodemographic factors and psychiatric disorders with type of school-based mental health services received by youth. Journal of Adolescent Health, 67(3), 392–400. https://doi.org/10.1016/j.jadohealth.2020.02.016

33.

Grindal

Schifter

L. A.

Schwartz

Hehir

. (2019). Racial differences in special education identification and placement: Evidence across three states. Harvard Educational Review, 89(4), 525–553. https://doi.org/10.17763/1943-5045-89.4.525

34.

Halbesleben

J. R. B.

Neveu

J.-P.

Paustian- Underdahl

S. C.

Westman

. (2014). Getting to the “COR”: Understanding the role of resources in conservation of resources theory. Journal of Management, 40(5), 1334–1364. https://doi.org/10.1177/0149206314527130

35.

Jackson

Makarin

. (2018). Can online off-the-shelf lessons improve student outcomes? Evidence from a field experiment. American Economic Journal: Economic Policy, 10(3), 226–254. https://doi.org/10.1257/pol.20170211

36.

Johnson

S. M.

Kraft

M. A.

Papay

J. P

. (2012). How context matters in high-need schools: The effects of teachers’ working conditions on their professional satisfaction and their students’ achievement. Teachers College Record, 114(10), 1–40.

37.

Jones

Winters

(2020). Are two teachers better than one? The effect of co-teaching on students with and without disabilities. Working paper. http://sites.bu.edu/voise/files/2020/06/Effect-of-Co-Teaching.pdf

38.

Jones

Youngs

. (2012). Attitudes and affect: Daily emotions and their association with the commitment and burnout of beginning teachers. Teachers College Record, 114(2), 1–36. https://eric.ed.gov/?id=EJ998988

39.

Kalyanpur

Harry

Skrtic

. (2000). Equity and advocacy expectations of culturally diverse families’ participation in special education. . International Journal of Disability, Development and Education, 47(2), 119–136. https://doi.org/10.1080/713671106

40.

Kennedy

. (2016). How does professional development improve teaching? Review of Educational Research, 86(4), 945–980. https://doi.org/10.3102/0034654315626800

41.

Kincaid

A. P.

Sullivan

A. L

. (2017). Parsing the relations of race and socioeconomic status in special education disproportionality. Remedial and Special Education, 38(3), 159–170. https://doi.org/10.1177/0741932516671199

42.

Klingner

J. K.

Vaughn

. (2002). The changing roles and responsibilities of an LD specialist. Learning Disability Quarterly, 25(1), 19–31. https://doi.org/10.2307/1511188

43.

Kostyo

Cardichon

Darling-Hammond

(2018). Making ESSA’s Equity promise real: State strategies to close the opportunity gap. Learning Policy Institute.

44.

Kraft

M. A.

Papay

J. P

. (2014). Can professional environments in schools promote teacher development? Explaining heterogeneity in returns to teaching experience. Educational Evaluation and Policy Analysis, 36(4), 476–500. https://doi.org/10.3102/0162373713519496

45.

Mason-Williams

. (2015). Unequal opportunities: A profile of the distribution of special education teachers. Exceptional Children, 81(2), 247–262. https://doi.org/10.1177/0014402914551737

46.

Mason-Williams

Bettini

Peyton

Harvey

Rosenberg

Sindelar

P. T

. (2020). Rethinking shortages in special education: Making good on the promise of an equal opportunity for students with disabilities. Teacher Education and Special Education, 43(1), 45–62. https://doi.org/10.1177/0888406419880352

47.

Molloy

L. E.

Moore

J. E.

Trail

Van Epps

J. J.

Hopfer

(2013). Understanding real-world implementation quality and “active ingredients” of PBIS. Prevention Science, 14, 593–605. https://doi.org/10.1007/s11121-012-0343-9

48.

Newmann

Smith

Allensworth

Bryk

. (2001). Instructional program coherence: What it is and why it should guide school improvement policy. Educational Evaluation and Policy Analysis, 23(4), 297–321. https://doi.org/10.3102/01623737023004297

49.

Nguyen

T. D.

Pham

L. D.

Crouch

Springer

M. G

. (2020). The correlates of teacher turnover: An updated and expanded meta-analysis of the literature. Educational Research Review, 31, 100355. https://doi.org/10.1016/j.edurev.2020.100355

50.

Nguyen

T. D.

Redding

. (2018). Changes in the demographics, qualifications, and turnover of American STEM teachers, 1988–2012. AERA Open, 4(3), 1–13. https://doi.org/10.1177/2332858418802790

51.

O’Brien

K. M.

Brunsting

N. C.

Bettini

Cumming

M. M.

Ragunathan

Sutton

. (2019). Special educators’ working conditions in self-contained settings for students with emotional or behavioral disorders: A descriptive analysis. Exceptional Children, 86(1), 40–57. https://doi.org/10.1177/0014402919868946

52.

Office of Special Education Programs (2019). <41st annual report to congress on the implementation of IDEA. https://www2.ed.gov/about/reports/annual/osep/index.html.

53.

Peyton

van Dijk

Mason-Williams

Kramer

(2020). Teaching to the money: Impact of TEACH grant on education degree production. Working Paper.

54.

Prather-Jones

. (2011). “Some people aren’t cut out for it”: The role of personality factors in the careers of teachers of students with EBD. Remedial and Special Education, 32(3), 179–191. https://doi.org/10.1177/0741932510362195

55.

Rockoff

. (2004). The impact of individual teachers on student achievement: Evidence from panel data. American Economic Review, 94(2), 247–252. www-jstor-org.web.bisu.edu.cn/stable/3592891

56.

Ronfeldt

Farmer

S. O.

McQueen

Grissom

J. A

. (2015). Teacher collaboration in instructional teams and student achievement. American Educational Research Journal, 52(3), 475–514. https://doi.org/10.3102/0002831215585562

57.

Rothstein

. (2015). The Americans with disabilities Act and higher education 25 years later: An update on the history and current disability discrimination issues for higher education. Journal of College and University Law, 41(3), 531. http://ssrn.com/abstract=2629306

58.

Rozalski

Stewart

Miller

. (2010). How to determine the least restrictive environment for students with disabilities. Exceptionality, 18(3), 151–163. https://doi.org/10.1080/09362835.2010.491991

59.

Sayeski

K. L.

Bateman

D. F.

Yell

M. L

. (2019). Re-envisioning teacher preparation in an era of Endrew F.: Instruction over access. Intervention in School and Clinic, 54(5), 264–271. https://doi.org/10.1177/1053451218819157

60.

Scruggs

T. E.

Mastropieri

M. A.

McDuffie

K. A

. (2007). Co-teaching in inclusive classrooms: A metasynthesis of qualitative research. Exceptional Children, 73(4), 392–416. https://doi.org/10.1177/001440290707300401

61.

Simon

N. S.

Johnson

S. M

. (2015). Teacher turnover in high-poverty schools: What we know and can do. Teachers College Record, 117(3), 1–36. https://eric.ed.gov/?id=EJ1056722

62.

Sindelar

P. T.

Shearer

D. K.

Yendol-Hoppey

Liebert

T. W

. (2006). The sustainability of inclusive school reform. Exceptional Children, 72(3), 317–331. https://doi.org/10.1177/001440290607200304

63.

Siuty

M. B.

Leko

M. M.

Knackstedt

K. M

. (2018). Unraveling the role of curriculum in teacher decision making. Teacher Education and Special Education, 41(1), 39–57. https://doi.org/10.1177/0888406416683230

64.

Tatto

M. T.

Savage

Liao

Marshall

S. L.

Goldblatt

Contreras

L. M

. (2016). The emergence of high-stakes accountability policies in teacher preparation: An examination of the US Department of Education’s Proposed regulations. Education Policy Analysis Archives, 24, 21. https://doi.org/10.14507/epaa.24.2322

65.

Theobald

Goldhaber

Naito

Stein

(2020). The special education teacher pipeline: Teacher preparation, workforce entry, and retention. Working Paper No. 231-0220. National Center for Analysis of Longitudinal Data in Education Research (CALDER). https://caldercenter.org/sites/default/files/CALDER%20WP%20231-0220_0.pdf

66.

Vannest

K. J.

Hagan-Burke

. (2010). Teacher time use in special education. Remedial and Special Education, 31(2), 126–142. https://doi.org/10.1177/0741932508327459

67.

Wexler

Kearns

Lemons

C. J.

Mitchell

Clancy

Davidson

Wei

. (2018). Reading comprehension and co-teaching practices in middle school English language arts classrooms. Exceptional Children, 84(4), 384–402. https://doi.org/10.1177/0014402918771543

68.

Youngs

Jones

Low

. (2011). How beginning special and general education elementary teachers negotiate role expectations and access professional resources. Teachers College Record, 113(7), 1506–1540. https://doi.org/10.1177/016146811111300704

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

0.42 MB

	Wave: 2000			Wave: 2012			Wave: 2016
	Low FRPL	High FRPL	Diff	Low FRPL	High FRPL	Diff	Low FRPL	High FRPL	Diff
Demands
Self-contained class	0.31	0.43	0.12**	0.24	0.36	0.12**	0.25	0.34	0.09**
Co-teaching	0.10	0.06	−0.04**	0.16	0.17	0.01	0.14	0.12	−0.02
Resource delivery model	0.41	0.37	−0.04	0.43	0.32	−0.1**	0.45	0.41	−0.03 +
Total hours in school	47.99	47.91	−0.09	49.65	50.18	0.52	51.23	50.73	−0.51
Duties interfere	2.04	2.04	0.00	2.06	2.03	−0.03	2.12	2.08	−0.03
Number of grades taught	3.12	3.23	0.11	3.42	3.32	−0.1	3.32	3.43	0.11 +
Case load size	16.45	17.01	0.57	20.70	19.24	−1.46	22.72	20.55	−2.17*
Student issue (std)	−0.24	−0.02	0.2**	−0.28	−0.03	0.26**	−0.24	0.12	0.36**
Internal Resources
No certification	0.05	0.1	0.05**	0.00	0.00	0.00	0.02	0.02	0.00
Experience	13.15	12.7	−0.46	13.64	13.04	−0.6	13.42	12.73	−0.69 +
SPED field of study	0.75	0.68	−0.07**	0.7	0.63	−0.06 +	0.7	0.65	−0.05 +
School-Based Resources
Teacher coop.	2.07	1.98	-.09	2.20	2.12	-.08*	2.23	2.15	−0.08*
Admin. support	2.15	2.07	−0.08	2.33	2.25	−0.08*	2.36	2.31	−0.05
Adequate materials	1.95	1.76	−0.19**	2.20	2.00	−0.20**	2.12	2.00	−0.12**
Observations	4,100			4,240			2,720