Abstract
The formula used to allocate federal funding to states for special education is one of IDEA's most critical components. The formula serves as the primary mechanism for dividing available federal dollars among states and represents policy makers’ intent to equalize educational opportunities for students with disabilities nationwide. In this study, we evaluate the distribution of IDEA Part B funding in the wake of changes to the formula that were instituted at the law's 1997 reauthorization. We find that the revised formula generates large and concerning disparities among states in federal special education dollars. On average, states with proportionally larger populations of children and children living in poverty, children identified for special education, and non-White and Black children receive fewer federal dollars per capita.
Policy makers are frequently criticized for the limited role the federal government plays in funding the Individuals With Disabilities Education Act (IDEA, 2004). Drawing the most attention has been the federal government's failure to “fully fund” Part B of IDEA, which provides federal dollars to states to pay for special education for children with disabilities (National Council on Disability, 2018). IDEA currently authorizes federal appropriations of up to 40% of average per-pupil spending nationwide—and yet, appropriations have never reached this target. Considerably less attention, however, has been paid to whether IDEA Part B appropriations are distributed to states in an equitable and efficient manner.
The formula that the federal government uses to allocate IDEA funding to states is one of the law's most critical components. It not only is the mechanism for determining the amount of funding state and local education agencies (LEAs) receive each year but also represents policy makers’ priorities for who should receive more and less federal aid for special education programs (Dragoo, 2019). At IDEA's inception, policy makers intended federal funding to be divided among states proportionally according to differences in the demand for special education services (U.S. Congress Senate Committee on Labor and Public Welfare, 1975). The formula was substantially revised at IDEA's 1997 reauthorization to address concerns about the role federal funding might play in incentivizing states to overidentify children for special education (U.S. Congress House of Representatives Committee on Education and the Workforce, 1997; U.S. Congress Senate Committee on Labor and Human Resources, 1997) and now allocates all new federal special education funding based on states’ populations of school-age students and children experiencing poverty.
There are growing concerns that the changes to the funding formula put in place at the 1997 IDEA reauthorization introduced inequities in the distribution of federal funding among states (Harr & Parrish, 2005; Kolbe et al., 2022; McCann, 2014). In this study, we evaluate whether the existing formula equitably distributes funding among states. To do so, we use (a) the criteria that were articulated by policy makers when crafting IDEA's funding formula, both prior to and after the new formula went into effect, and (b) standards used by policy makers to develop and the courts to evaluate contemporary K–12 school finance policies. We find that changes to the formula created new disparities in federal funding among states that systematically disadvantage states with larger populations of K–12 students and children experiencing poverty, students receiving special education, and non-White and Black students. Such differences do not reflect policy makers’ intent to ensure at least parity in IDEA Part B funding among states and are out of step with school funding principles that call for distributing funding in ways that adjust for the demand for special education services, student need, and educational costs.
Policy Context
Federal Funding for Special Education
Starting with the Education for All Handicapped Children Act of 1975 and including its later reauthorizations as IDEA, the federal government has operated a permanent funding program that provides categorical aid to states to offset the additional cost of providing special education and related services to children with disabilities. Of the appropriations authorized by IDEA, the largest funding program is Section 611 of Part B that awards grants to states to pay for a portion of the excess cost of providing special education and related services to children with disabilities ages 3 to 21. 1 Over time, IDEA Part B has evolved into one of the federal government's primary “funding statutes” for public education (Mead, 2017, p. 21). For fiscal year (FY) 2022, federal policy makers appropriated $15.5 billion to pay a portion of the costs of educating nearly 7 million children with disabilities nationwide (U.S. Department of Education, 2021).
When IDEA was first authorized, the federal government determined that it would fund 40% of this excess cost (Dragoo, 2019). Yet historically, IDEA appropriations have fallen well below this goal. Typically, the percentage has not exceeded 20%, and in FY 2020, the federal share was 13.2%. The smallest gap occurred when states received stimulus funds to address the Great Recession in FY 2009, at which time funding reached 35% of the national average per-pupil spending (Dragoo, 2019; Zembar, 2021). As states and LEAs continue to cover the difference in cost with their own revenues, policy makers, administrators, and stakeholders have increasingly called for “full funding” of IDEA by the federal government (e.g., National Council on Disability, 2018).
IDEA was designed to provide states and LEAs with a framework for equalizing educational opportunities for school-age children with disabilities nationwide (20 U.S.C. §1400[c][1]). Broadly, IDEA operationalizes the concept of equal educational opportunity in its core entitlements for students with disabilities—access to a free and appropriate public education (FAPE) in the least restrictive environment (Mclaughlin, 2010). These goals are reflected in the provisions articulated by IDEA for establishing an individualized educational program (IEP) for each child with a disability who is found eligible for special education. In exchange for federal funding, states agree to implement the law's detailed procedural requirements associated with the IEP to ensure that each child is treated equitably (Dragoo, 2019). States must not only oversee educational policy and practice by local educators but also pass through the majority of federal IDEA Part B funding to LEAs.
When crafting the law, policy makers also recognized that an equitable distribution of IDEA funding among states was “essential for the Federal government to meet its responsibility to provide an equal educational opportunity” (20 U.S.C. §1400[c][7]). The primary policy mechanism for equalizing aid among states is the formula articulated in federal statute and regulations that dictates how annual IDEA appropriations are distributed to states and LEAs. Congress initially operationalized its intent for equity in distributing IDEA funding by implementing a formula that was intended to provide states with different total grant amounts according to cross-state differences in the counts of students identified for special education, but with the assumption that the formula would provide states with roughly equal amounts of funding per student receiving special education among all states (Congressional Record, 17478, 1972; U.S. Congress Senate Committee on Labor and Public Welfare, 1975).
Calculating State IDEA Part B Grant Amounts
The way in which the IDEA Part B formula accounts for differences among states in the demand for special education services has changed over time. At the law's inception, policy makers were concerned that states may be either unable or unwilling to ensure children with disabilities access to FAPE (20 U.S.C. §1400[c][1–3]). The initial formula reflected these concerns, allocating federal dollars based on the number of students identified for special education. In other words, a state could expect its total federal grant to increase with growth in its share of students identified for special education (Congressional Record, 17478, 1972; U.S. Congress Senate Committee on Labor and Public Welfare, 1975). In turn, states were obligated to distribute funding to their LEAs using the same formula.
As part of the 1997 IDEA reauthorization, policy makers revised the formula with the intent of breaking the link between the number of children a state identified for special education and federal funding. This was done in response to concerns about (a) growth in the number of children identified for special education and the corresponding increase in demand for federal special education funding and (b) whether the existing approach to calculating federal aid motivated educators to “overidentify” children for special education, particularly children from minoritized populations and who attended urban schools (U.S. Congress House of Representatives Committee on Education and the Workforce, 1997; U.S. Congress Senate Committee on Labor and Human Resources, 1997).
Policy makers added a “census-based” component to the formula's calculations that distributed IDEA Part B funding to states using both a population and a poverty calculation. For the census-based population calculation, 85% of new federal appropriations are allocated to states based on their share of the national population of children ages 3 to 21 (Dragoo, 2019). The other 15% is allocated according to states’ shares of the national population of children experiencing poverty (Dragoo, 2019). The population and poverty calculations were intended to serve as a proxy for cross-state differences in the prevalence of childhood disability in the population—but without explicit ties to state and local decisions about the eligibility of children for special education services (U.S. Congress House of Representatives Committee on Education and the Workforce, 1997; U.S. Congress Senate Committee on Labor and Human Resources, 1997). In its most straightforward application, states with larger populations of school-age children and children experiencing poverty would receive larger total grants, whereas all states would receive about the same amount of funding per pupil (Parrish et al., 2015).
To ensure that states did not see a reduction in their total IDEA Part B grant amounts in the wake of the formula's changes, Congress guaranteed states, in perpetuity, a minimum base amount equivalent to the total grant amount they received for FY 1999 (in nominal dollars). Only new appropriations, exceeding the total FY 1999 IDEA Part B appropriation, are allocated using the new census-based population-poverty calculation (85/15). As a result, a state's grant is the sum of three factors: (a) the FY 1999 base year grant (in nominal dollars), (b) its share of new federal appropriations based on population, and (c) its share of new appropriations based on child poverty (20 U.S.C. §1411[e]). Policy makers also put in place other provisions intended to stabilize the amount of funding states receive across years. Once a state's initial grant is calculated, the amount may be adjusted for a minimum or maximum award. The minimum is determined by comparing four alternative calculations that evaluate the grant amount based on prior-year levels and different assumptions about its share of annual IDEA Part B appropriations. The maximum award limits the year-to-year increase in total funding a state can receive.
The new multistep formula and the corresponding minimum and maximum grant calculations went into effect for FY 2000 allocations, the first year following the 1997 reauthorization that IDEA Part B appropriations exceeded $4.9 billion. The formula was largely unchanged at IDEA's 2004 reauthorization and remains current policy.
Distribution of IDEA Part B Funding Among States
Changes to the IDEA Part B formula put in place with the 1997 reauthorization altered the distribution of federal dollars among states in the years immediately following the new formula's implementation. By FY 2003, over half (51.4%) of Part B funding was allocated using the census-based calculation, and with this shift came increased variation among states in Part B funding per special education student (Harr & Parrish, 2005). Although there was almost no difference among states in federal revenues per special education student for FY 1999, by FY 2003, Part B dollars per student receiving special education in the state getting the most per student were 43% greater than in the state receiving the least ($1,503 vs. $1,051; Harr & Parrish, 2005). The emerging differences in state grant amounts per student were attributed to the fact that the formula allocated new federal dollars using states’ population and poverty counts.
Using data from a national survey conducted with states and districts, McCann (2014) revisited the question of the distribution of IDEA appropriations among states and districts and found substantial variation in federal IDEA Part B funding per pupil (i.e., total enrollment). More recently, Kolbe et al. (2022) evaluated policy proposals that would increase federal appropriations for IDEA to determine how new funding would be distributed among states. They found that the existing formula would not distribute new IDEA Part B appropriations in ways that reflect cross-state differences in the demand for special education services. Rather, for FY 2020, the existing formula systematically disadvantaged states with larger populations, along with states with larger shares of children experiencing poverty, children with disabilities, and non-White children. Policy simulations also showed how using the existing formula to allocate proposed new funding for IDEA Part B would worsen existing disparities among states in federal funding both per pupil and per student receiving special education.
Study Overview
The purpose of this study is to evaluate changes in the distribution of federal funding that occurred during the 20-year period after the new formula went into effect, particularly whether formula changes systematically impacted certain states and student populations. Specifically,
In what ways did the distribution of IDEA Part B funding among states change after FY 2000, when the new federal formula went into effect? To what extent do contemporary differences among states in IDEA Part B funding reflect cross-state differences in the demand for special education services and student disadvantage?
Our evaluation is grounded in contemporary K–12 school funding principles, where the central assumption is that education finance systems should provide children with equal educational opportunities, regardless of where they live (Baker, 2018; Baker et al., 2007). These principles are evident in state education policies and court cases that seek to (a) mitigate the relationship between where a child lives and attends school, especially with respect to differences in local wealth, and (b) provide compensatory funding to account for differences in the cost of equalizing education opportunities for all students. Similar logic can be extended to federal funding for state and local special education programs, where the broad policy objective is to equalize opportunities for children with disabilities by differentiating aid according to population-based differences in the demand for special education and related services among states (20 U.S.C. §1400[c][7]).
By design, the existing policy framework and calculations used to allocate IDEA Part B funding are intended to generate differences among states in the total grant amount. Consistent with the law's intent, prior to FY 2000, the total federal grant amount would be higher in states that, because of the number of children eligible for special education, face a higher total cost of implementing the law's provisions for identifying and ensuring FAPE for children with disabilities. Following the formula's revisions, we would expect that the total federal grant amount would be higher in states with more school-age children and children experiencing poverty. However, although the total amount of funding should vary among states, at a minimum there is an expectation for nominal parity across states in funding—that is, states should receive roughly equal federal dollars per pupil. We would also expect the revised formula to provide states with roughly equal federal dollars per student receiving special education if states’ student population and child poverty counts work as proxies for the number of children needing special education in a state, as policy makers intended. As a first step in our analysis, we evaluate the distribution of federal IDEA Part B funding among states to determine to what extent states received different levels of funding per capita and in what ways the distribution of federal funding has changed since FY 2000, when the new formula went into effect.
Simply allocating the same amount of funding per capita—either per student receiving special education or per pupil—may still be inequitable. Contemporary K–12 school finance research also shows that a horizontally equitable distribution of funding (i.e., equal treatment of equals; Berne & Stiefel, 1984) may be insufficient to equalize educational opportunity. Instead, to equalize opportunity, resources should be differentiated vertically, with respect to different student, school, and community characteristics that are proxies for differences in need or cost of providing an adequate education (Baker, 2018; U.S. Department of Education, 2021). Equity concerns arise when opportunities to learn are further skewed in ways that disadvantage children from certain backgrounds or locations (Berne & Stiefel, 1984; Rodriguez, 2004). Accordingly, there may be good reason for federal IDEA Part B funding to vary among states, including factors that increase disadvantage and differences in need for special education services. As a second step in our analysis, we consider whether the existing formula distributed IDEA Part B appropriations progressively. A progressive distribution would shift more federal funding per capita to states with greater demand for special education services, student economic disadvantage, and historic and current racial marginalization and discrimination (Baker, 2018; Rodriguez, 2004; Toutkoushian & Michael, 2007).
Method
Data
Information about states’ IDEA Part B grant amounts was obtained from the U.S. Department of Education state tables compiled by the Department of Education's Budget Services (U.S. Department of Education, n.d.). These data include detailed information on the federal grant aid allocations to state education agencies, including each state's total allocation for IDEA Part B (611) dollars. We merged these data with other sources that provide additional descriptive information for each state's educational context, including (a) student enrollment and demographics from the Department of Education's Common Core of Data (CCD; National Center for Education Statistics, n.d.); (b) the number of students with IEPs in a state from the Office of Special Education Programs' IDEA Part B child count data (U.S. Department of Education, n.d.); (c) the population of children experiencing poverty in a state, as reported by the U.S. Census Bureau (U.S. Census Bureau, n.d.); 2 and (d) the head count of children (ages 3–21) living in a state from the U.S. Census Bureau's annual estimates from the decennial census. 3 We constructed a data set that included annual federal aid allocations and other state characteristics (from all sources) for FYs 1999 to 2021, the period immediately prior to and then following when the new formula was implemented.
For comparison purposes, we standardized a state's total IDEA Part B grant amounts two ways: (a) grant dollars per student receiving special education—that is, total IDEA Part B grant award divided by a state's special education child count—and (b) grant dollars per pupil—that is, the total IDEA Part B grant award divided by a state's total average daily membership (student head count). The first measure allows us to consider whether nominal parity exists among states with respect to the amount of federal funding available to support students receiving special education in a state, consistent with policy makers’ original intent (as expressed in IDEA, prior to the 1997 reauthorization). Alternatively, the second measure—that is, state grant dollars per pupil—is not directly related to the number of children receiving special education services and allows us to consider the extent to which the new population-poverty calculations introduced in the revised formula provide nominal parity among states based on student population.
Analytic Strategy
Our descriptive analyses follow the concepts and methods used in K–12 education finance policy research to evaluate resource differences across and within states—specifically to (a) evaluate the variation in federal grant aid and (b) identify systematic differences among states according to what can be explained by relevant need and other factors (Baker, 2018; Berne & Stiefel, 1984; Downes & Stiefel, 2015; Kolbe et al., 2022; Needham & Houck, 2019).
Distribution of IDEA funding among states
First, we illustrate how the share of federal appropriations allocated by different parts of the formula—the FY 1999 base amount and the census-based population-poverty calculation—changed between FYs 1999 and 2021 and juxtapose this trend with how state grant amounts per pupil and per student receiving special education changed over time.
Second, we evaluate the extent of variation in IDEA Part B funding among states using two commonly used measures of horizontal equity in K–12 school spending: (a) the coefficient of variation (CV) and (b) the McLoone Index (Berne & Stiefel, 1984; Toutkoushian & Michael, 2007). Horizontal equity measures treat all states as if they have a similar demand for special education (i.e., equal treatment of equals) and describe the unconditional variation in IDEA Part B grant amounts among states (Toutkoushian & Michael, 2007).
The CV illustrates the extent to which states received similar IDEA Part B funding amounts per pupil and per student receiving special education. CV values near zero suggest parity in funding among states in federal funding, whereas larger values signal greater disparity. The U.S. Department of Education uses a measure like the CV to calculate the equity factor in determining states’ Title I Education Finance Incentive Grant amounts (Sonnenberg, 2016). A CV of at least 0.10 is the standard for an equitable distribution of resources (Kelly, 2015; Odden & Picus, 2004). The McLoone Index assumes that if students are rank ordered according to the amount of state-level IDEA Part B funding received, perfect equity would be achieved if every student received at least as much as the middle of the distribution. It is calculated as the ratio of the funding allocated to students below the median to the funding needed to raise all students to the median allocation. Values have a lower bound of zero, and higher values suggest a more equitable distribution of funding among states. The McLoone Index not only tells us about the extent of disparity; it also supports calculating the dollars necessary to ensure that all states have at least as much IDEA funding (per pupil or per student receiving special education) as the median state. This supplemental calculation is particularly relevant for evaluating IDEA funding, where the policy discussion often focuses on increasing federal funding for special education programs. It answers the question about how much additional funding is needed to ensure all states are at least funded at the median state per-capita grant amount. A McLoone Index value greater than 95% is the standard for an equitable distribution of resources (Kelly, 2015; Odden & Picus, 2004).
Evaluating the distribution of IDEA funding among states
Third, we examine whether differences in IDEA Part B grant amounts vary according to factors identified by policy makers as proxies for demand for special education services and other indicators of student need and discrimination. We start by comparing quartiles, dividing the data into four equal sizes based on several metrics explained later. We focus on the top and bottom quartiles to summarize the difference between the top 25% of states (largest values) and the bottom 25% of states (smallest values). Other quartiles can be compared, but for this analysis, our focus is illustrating the difference between the top and bottom states. Initially, we compare average funding per pupil and per student receiving special education for states in the top and bottom quartiles, where states are ranked according to the fraction of the total U.S. population of (a) children ages 3 to 21, (b) children experiencing poverty, and (c) the percentage of school-age children identified for special education. Additionally, we compare states with the largest and smallest concentrations of non-White and Black school-age children. There are long-standing concerns about the racialization of K–12 education funding, particularly that existing funding formulas provide fewer dollars to places that serve large concentrations of minoritized students (Baker et al., 2022; Hammond, 2004; Rothbart, 2020). In this study we explore whether the existing IDEA funding formula similarly discriminates in distributing federal aid to states for special education programs. For our analysis, we use the same measure for child poverty that is used in the existing formula's poverty calculation (i.e., U.S. Census Bureau's Small Area Income Poverty Estimates). 4 Data from the CCD were used to identify the number of students enrolled in K–12 public schools in a state who identify as White and Black and to calculate a state's percentage of students who were non-White (i.e., all students not identified as White) and Black.
We then used regression analysis to evaluate vertical equity in the distribution of federal funding among states (Berne & Stiefel, 1984; Downes & Pogue, 2002; Toutkoushian & Michael, 2007). We relate states’ IDEA Part B funding (per pupil and per student receiving special education) to the vector of state population characteristics considered in the quartile analysis. A positive correlation between state grant amounts suggests a progressive distribution in federal funding among states on identified factors. Dependent variables are states’ FY 1999 and FY 2020 per-pupil spending (columns 1 and 2) and per-student-receiving-special-education allocation (columns 3 and 4). The independent variables are a state's percentage (a) of non-White children, (b) school-age children identified for special education, (c) children experiencing poverty, and (d) children ages 3 to 21.
Results
Distribution of IDEA Funding Among States
Congress increased the federal IDEA Part B appropriation from about $4.8 billion in FY 1999 to $15.2 billion in FY 2021—a 264% increase in nominal dollars and a 123% increase in real dollars over a 23-year period. The largest increase in funding during this period occurring in FY 2021and FY 2022 with nearly $2.58 billion in new appropriations, which in FY 2021 included federal COVID emergency relief funding. By design, as federal appropriations for IDEA Part B increased, more dollars were allocated to states using the formula's census-based component. For FY 2000, the first year the new formula went into effect, just 13.7% of federal funding was allocated using the census-based population-poverty calculation, and the remainder went toward states’ FY 1999 base amounts. By FY 2021, however, 72.8% of funding was allocated using the census-based population-poverty calculation, and just 27.2% of the total appropriations went toward funding states’ FY 1999 base amounts (Figure 1).
Percentage of Individuals With Disabilities Education Act Part B appropriations allocated to states using census-based calculation and states’ fiscal year 1999 base amount.
Increased federal appropriations translated into larger average state grant amounts, both per pupil and per student receiving special education (Table 1). Between FY 1999 and FY 2021, nationally, the average per-pupil grant increased 249% ($91 to $318) and 210% per student receiving special education ($783 to $2,489). However, not all states benefited equally from increased federal appropriations.
IDEA Part B State Grant Amounts per Pupil and per Student Receiving Special Education (FY 1999 and FY 2021).
Source. Authors’ calculations using U.S. Department of Education (n.d.), National Center for Education Statistics (n.d.), and Office of Special Education Programs (n.d.).
Note. IDEA = Individuals With Disabilities Education Act; FY = fiscal year.
States are listed in descending order according to the percentage change in per-pupil grant amount between FY 1999 and FY 2021.
Differences between the states receiving the most and least federal funding per pupil and per student receiving special education also grew over time as federal appropriations for IDEA Part B increased (Figures 2 and 3).
In FY 1999, the difference between the states at the top and bottom of the distribution was just $47 per pupil and $582 per student receiving special education. However, in the 20 years since the new formula went into effect, the gap widened to $252 per pupil and $1,396 per student receiving special education (FY 2021)—an increase of 436% and 139%, respectively.
Individuals With Disabilities Education Act Part B allocations to states per pupil, fiscal years 1999 to 2021 (nominal $).
Changes to the CV provide further insight into variability among state IDEA Part B grants over time (Figure 4). For FY 2000, the CV was 3.6% for state grant amounts per student receiving special education and 11.6% for state grants per pupil. The comparatively smaller CV for state grant funding per student receiving special education reflects the fact that initially the new formula allocated most appropriations according to states’ FY 1999 base amount, which was tied to a state's special education child count for the prior academic year. The larger CV describing variability in state grant dollars on a per-pupil basis shows how in the formula's first year of implementation there was far less nominal parity among states in IDEA Part B funding when considered in terms of states’ overall student head counts.
Individuals With Disabilities Education Act Part B allocations to states per student receiving special education, fiscal years 1999 to 2021 (nominal $).
Coefficient of variation for state Individuals With Disabilities Education Act Part B grants, per pupil and per student receiving special education (fiscal years 1999–2021).
Variability in state grant amounts increased over time with additional new federal appropriations for IDEA Part B (Figure 3). Early on, as the formula allocated more dollars using the census-based calculation, the CV for state grants per student receiving special education steadily increased. For FY 2010, the CV for state grants per student receiving special education nearly tripled (11.8%) but increased just 2 percentage points for state grants per pupil. At that time, nearly half of federal appropriations were allocated according to states’ child population and poverty counts, and the FY 1999 base amount began to operate more as a fixed grant rather than a variable grant that reflected differences among states in special education child count. By FY 2021, the CVs for state grants per student receiving special education and per pupil were at their highest levels since the new formula went into effect in FY 2000: 16.3% for grant dollars per pupil and 14.2% for grant dollars per student receiving special education. CVs of this magnitude are on par with levels that raise concern when evaluating fiscal equity in K–12 education finance (Baker et al., 2007).
The McLoone Index provides information about the bottom of the distribution of state grant amounts compared with the middle of the distribution (median). We find that for FY 1999, total IDEA Part B appropriations were 97% of what was needed to raise the allocation per student receiving special education to the median allocation nationally and 93% of what was needed for the per-pupil allocation. In the past two decades, the indices fell to 93% per student receiving special education and 87% per pupil, suggesting that $943,113,216 would be necessary to bring all states’ grant amounts to the median per-pupil amount and $509,338,624 to bring all states to the median amount per student receiving special education.
Evaluating the Distribution of IDEA Funding Among States
Differences among states in IDEA Part B grant amounts are not inherently inequitable. A progressive approach to allocating federal aid to states would direct more resources to states with higher costs of educating children with disabilities because of either prevalence in the population, extent of need and demand for services, or local prices. However, in the case of variability in IDEA Part B grants to states, we find evidence to the contrary.
Prior to the formula change, states largely received similar funding per student receiving special education (Table 2).
Average IDEA Part B Grant Amounts for States in Top and Bottom Subgroup Quartiles (FY 1999 and FY 2021).
Source. Authors calculations using U.S. Department of Education (n.d.), National Center for Education Statistics (n.d.), Office of Special Education Programs (n.d.), U.S. Census Bureau Population Division (n.d.), and U.S. Census Bureau Small Area Income Poverty Estimates Program (n.d.).
Note. IDEA = Individuals With Disabilities Education Act; FY = fiscal year; Q = quartile.
However, by FY 2021, there were substantial differences among states in federal IDEA Part B grant amounts.
For instance, states in the top quartile for the share of children experiencing poverty received on average 16% less per pupil and 10% less per student receiving special education than states with grant amounts in the bottom quartile, with the smallest fractions of children experiencing poverty (Table 2). Similarly, states with the largest populations of school-age children received IDEA Part B grants of 16% less per pupil and 12% less per student receiving special education than the smallest states. The difference is even more pronounced when we compare states that saw the largest and smallest gains in child population in the past two decades. For FY 2021, states with the largest increases in child population since the formula change received about 28% fewer federal grant dollars per pupil and 9% fewer dollars per student receiving special education than their counterparts that saw the smallest increases (or even decreasing) child populations since FY 1999 (Table 3). Interestingly, states with the largest increases of children experiencing poverty received 6% fewer dollars per pupil than states with the smallest increases but 5% more dollars per student receiving special education.
Average IDEA Part B Grant Amounts for States in Top and Bottom Subgroup Quartiles for Population Changes Between FY 1999 and FY 2021.
Source. Authors’ calculations using U.S. Department of Education (n.d.), National Center for Education Statistics (n.d.), U.S. Census Bureau Population Division (n.d.), and U.S. Census Bureau Small Area Income Poverty Estimates Program (n.d.).
Note. IDEA = Individuals With Disabilities Education Act; IEP = individualized education program; FY = fiscal year; Q = quartile.
The distribution of federal funding also systematically disadvantaged states with the largest shares of minoritized children by providing them with fewer IDEA Part B dollars per capita (Table 2). For FY 2021, per-pupil grant amounts for states with the largest shares of non-White and Black students were 17% and 11% lower (respectively) than states with the smallest shares of minoritized children. Grant amounts per student receiving special education were also lower in states with the greatest concentrations of minoritized children—a difference of 10% between states with the largest and smallest shares of non-White students and a 4% discrepancy for states with the largest and smallest shares of Black students.
Table 4 shows the relationships between state characteristics and the distribution of federal IDEA Part B grant dollars (per student receiving special education and per pupil) while simultaneously accounting for the complement of state characteristics examined already. The purpose of this analysis is to examine whether there were changes in the overall distribution of IDEA Part B funding among states between FY 1999 and FY 2021. Taken together, the full models affirm that the existing formula introduced new systematic differences in states’ grants. In FY 2021, states with larger percentages of children with disabilities received fewer grant dollars per student receiving special education, whereas places with higher percentages of children experiencing poverty received more dollars per student receiving special education. States with the largest populations of children ages 3 to 21, however, received substantially fewer dollars per student receiving special education. For example, a state with a child population of one standard deviation above the mean received, on average, almost $700 less per student receiving special education, and conversely, a state one standard deviation below the mean received $700 more—a $1,400 difference per student between states with the largest and smallest child populations.
Explaining Variation in State Grant Amounts per Pupil and per Student Receiving Special Education (FY 2021).
Note. Dependent variable for each column is per-pupil spending (columns 1 and 2) or per-student-receiving-special-education allocation (columns 3 and 4). The independent variables are the fraction of the total U.S. population (a) of non-White children (% non-White), (b) the percentage of school-age children identified for special education (% special education students), (c) children living in poverty (% child poverty), and (d) children ages 3 to 21 (% child population). Robust standard errors are reported. FY = fiscal year.
*p < .05. **p < .01. ***p < .001.
Discussion
Federal funding programs for public education are largely grounded in a “redistributive” or “equity” rationale (National Research Council, 1999, p. 259)—relying on national wealth to offset some of the differences in costs of educating students with diverse learning needs. In doing so, the intent is to mitigate differences in opportunities to learn that result from states and localities being either unwilling or unable to provide the additional funding necessary to meet their students’ needs (Gordon, 2016). The formulas contained in federal statute are the primary policy mechanism for distributing federal aid to accomplish these goals. In this study, we examined how the IDEA Part B (611) formula allocated federal funding to states since FY 2000, the period following the formula revisions put in place by the 1997 reauthorization.
We find that the existing formula currently generates large and concerning disparities in the distribution of federal special education dollars among states—to an extent that may work against the law's purpose of equalizing educational opportunities for students with disabilities across states.
Disparities among states in federal funding for special education programs—either per pupil or per student receiving special education—largely did not exist at the time the new formula went into effect. Rather, differences in state grant amounts grew over time in the wake of changes to the law that established a two-part funding system—with a fixed dollar grant tied to funding levels for FY 1999 and a new census-based system that allocated all new federal appropriations according to states’ child population and poverty counts. By fixing the FY 1999 base amount in time—neither increasing nor decreasing in nominal dollars from what each state received that year—nearly 40% of federal appropriations continue to be distributed based on cross-state differences in special education child counts that existing for the 1997 academic year. As the base funding amount proportionally decreased as a share of total funding and larger shares of federal dollars were allocated using the census-based calculation, variability in state grant amounts increased.
The formula change resulted in a fundamental shift in the distribution of funding among states, from one where there was similar funding for a child receiving special education, regardless of the state where they received services, to a system where the amount of funding available to serve students with disabilities was largely contingent on other state characteristics (e.g., child population and poverty) and in ways that systematically privileged and disadvantaged certain states.
Taken together, the study's findings show that distribution of federal IDEA funding among states that results from the existing formula fails to meet the criteria established by policy makers when crafting the law as well contemporary standards used by policy makers and the courts to develop and evaluate K–12 school funding programs.
Implications
Concerns over the distribution of federal IDEA funding among states are not new. That said, new policy proposals to significantly increase IDEA Part B appropriations—including recent efforts on the part of the Biden-Harris administration to “fully fund” IDEA—bring a new sense of urgency to creating a more equitable funding formula. Moving forward, achieving goals for more equitably allocating IDEA Part B dollars will require changes to the statutory formula used to calculate state grant allocations.
That said, although ensuring equal educational opportunities for children with disabilities is foundational to IDEA, current law is not explicit with respect to what constitutes an equitable distribution of federal funds. Going forward, policy makers who are interested in redesigning the formula will need to grapple with the question, “What is a fair distribution of federal IDEA Part B funding among states?” Findings from this study suggest that in answering this question, policy makers should consider the following:
Is the policy objective to equalize funding per student receiving special education or according to state population characteristics? If the goal is for federal funding to offset the additional expense incurred by states and localities in providing special education and related services to children with disabilities, equalizing funding among states on a per-student-receiving-special-education-services basis is aligned with this goal. Alternatively, if the goal is for IDEA funding to serve as compensatory aid, more generally, or to introduce flexibility in how this aid is used (e.g., for early intervention), then equalizing funding according to population characteristics (e.g., student head count or child poverty) is aligned with these goals. What does it mean to “equalize” funding? For instance, is the goal nominal parity per capita among states or a progressive distribution that provides more aid per capita to places with higher prices or differences in need among students? Currently, statute does not consider differences in educational costs and levels of need in distributing aid to states. Additionally, like nearly all other federal and state K–12 school funding, the formula used to distribute IDEA Part B funding also does not account for disparities in school resources that result from racialized segregation and discrimination.
Practically, answers to these questions should be “cost based”—that is, tied to established standards for the types and amounts of resources required to implement effective special education programs and to meet differing student needs (Baker, 2018; Kolbe, 2019, 2021). However, currently the field lacks reliable estimates for special education cost and the factors that account for differences in cost. The most recent estimates are now nearly 20 years old, predating significant shifts in education policy and best practices for serving students with disabilities. Estimating special education costs and identifying the factors that account for differences in costs are necessary requirements to move forward with developing a funding formula that allocates IDEA Part B funding equitably among states that likely serve children with disabilities with varying needs. For instance, special education costs may differ according to child age (e.g., early childhood vs. K–12), disability, and other characteristics (e.g., economic disadvantage).
The study's findings also raise questions about the formula's two-part distribution strategy that allocates aid first to states and then to districts within states. Although potential inequities in the distribution of resources between districts within states was not the focus of this study, past research suggests that requiring states to use the same formula to allocate federal funding to LEAs may further compound inequities in federal special education funding between and within states (McCann, 2014). A next step for future research should be to explore to what extent using the federal formula to allocate aid to LEAs creates similar funding inequities.