For other versions of this document, see http://wikileaks.org/wiki/CRS-RL31189 ------------------------------------------------------------------------------ Order Code RL31189 CRS Report for Congress Received through the CRS Web Individuals with Disabilities Education Act (IDEA): Identification and Misidentification of Children with Disabilities November 20, 2001 Richard N. Apling Specialist in Social Legislation Domestic Social Policy Division Congressional Research Service ~ The Library of Congress Individuals with Disabilities Education Act (IDEA): Identification and Misidentification of Children with Disabilities Summary This report discusses issues related to the identification and misidentification of children with disabilities. Misidentification can result from failing to identify those with disabilities, from identifying children with disabilities they do not have, and from delaying identifying children with disabilities. Congress has been, and continues to be, concerned about problems of misidentification. These concerns have been reflected in provisions of special education legislation -- most notably the Individuals with Disabilities Education Act (IDEA) -- that seek to ensure that children with disabilities are identified and receive special education and related services and that children not be incorrectly identified. For example, Congress has included provisions in IDEA funding formulas to discourage states from attempting to increase their shares of federal funding by over-identifying students with disabilities. Some disabilities, such as visual or hearing impairment, are relatively easy to recognize and are less susceptible to misidentification. The identification of other disabilities, for example learning disabilities, depend on more subjective measures and are more prone to misidentification. Students identified with these more "subjective" disabilities account for most of the school-age special education population. The distribution of disabilities is not uniform in the school-age population. Students with speech and language impairments predominate in the early grades. Students with learning disabilities account for most students with disabilities beginning in the middle elementary grades and continuing into high school. As students with milder disabilities graduate or drop out of school, those with more severe disabilities account for most of the population, although at these ages (19-22) most individuals with disabilities are no longer part of the elementary and secondary school population. Most of the concern about misidentification has centered on the perceived over- identification of African American students (especially African American males). While African Americans account for about 15% of the population age 6-21, they account for about 20% of students identified with disabilities. Although some portion of this higher rate might be explained by factors related to the occurrence of disabilities -- for example, greater poverty among African Americans -- such factors may not be the full explanation of the rates at which African American students are identified as mentally retarded (nearly 35% of all such students) and emotionally disturbed (more than 25% of these students). Other factors, such as teachers' subjective judgments, are likely to be involved. There is disagreement on what should be done about identification problems. Some argue that IDEA provides ample tools to deal with these problems, and only better implementation and more rigorous enforcement of current law are necessary. Others argue that IDEA needs to be "reformed," for example, by requiring states or school districts to devise corrective plans if there is evidence of over-identification. Contents Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Congressional Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 The 1975 Act . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Current Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Continuing Concerns About Misidentification . . . . . . . . . . . . . . . . . . . 3 Categorizing Disabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Numbers of Students Served by Special Education . . . . . . . . . . . . . . . . . . . . 6 Numbers of Children with Various Disabilities . . . . . . . . . . . . . . . . . . . . . . . 8 Gender Differences Among Students Identified as Disabled . . . . . . . . . . . . 11 Distribution of Disabilities by Race/Ethnicity . . . . . . . . . . . . . . . . . . . . . . . 12 Disability Rates Among Racial/Ethnic Groups . . . . . . . . . . . . . . . . . . 12 Racial/Ethnic Composition of Disability Groups . . . . . . . . . . . . . . . . 13 Possible Policy Alternatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 List of Figures Figure 1. Number of Children Served Under IDEA by Age (school year 1998-1999) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure 2. Number of Children with Various Disabilities by Age (school year 1998-1999) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Figure 3. Gender Breakdown of Three Disability Groups (elementary and secondary students, 1994 data) . . . . . . . . . . . . . . . . . . . . . 12 Figure 4. Composition of Population with Disabilities by Racial/Ethnic Groups (school year 1998-1999) . . . . . . . . . . . . . . . . . . . . . 13 Figure 5. Racial/Ethnic Composition of Total Population Ages 6-21, of All Students With Disabilities, and of Populations of Students with Selected Disabilities (school year 1998-1999) . . . . . . . . . . . . . . . . . . 15 Figure 6. Ratio of Percentage of A Disability Group to Percentage of Total Population for Racial/Ethnic Groups (school year 1998-1999) . . . . . 19 Individuals with Disabilities Education Act (IDEA): Identification and Misidentification of Children with Disabilities Overview For years, Congress has been concerned about the identification and misidentification of children with disabilities. These concerns have been reflected in provisions of special education legislation -- most notably the Individuals with Disabilities Education Act (IDEA) -- that seek to ensure that children with disabilities are identified and receive special education and related services and that children who are not disabled and do not require special education not be misidentified and mislabeled. Three general kinds of misidentification can occur: ! Under-identification -- that is, failing to identify children who have disabilities and need special education to succeed in school; ! Over-identification -- that is, classifying students with disabilities they do not have; and ! Late identification -- that is, delaying identification of students with disabilities until later in their schooling when special education services may be less effective. Misidentification has costs for students wrongly identified and for their families, as well as for the country as a whole. Failure to identify means that a child will not receive the services and rights that IDEA guarantees such children. Without services, the child with a disability may be less likely to succeed in school and as a result faces the possibility of diminished success in later life. The child who is wrongly classified as disabled may be stigmatized by being labeled, may be separated for large parts of the day from his or her peers, and as a result, may receive an inferior education. The child for whom identification (and therefore services) is delayed will miss the opportunity for early intervention, which, some argue, can prevent more severe problems from developing as the child matures and may prevent the child from requiring special education in later grades. Congressional Background The 1975 Act. Misidentification of students with disabilities has concerned Congress for decades. A central congressional concern that led to the enactment of the Education for All Handicapped Children Act of 1975 (P.L. 94-142 -- the previous nomenclature for IDEA) was that most disabled children were being CRS-2 inadequately served in public schools or were being excluded altogether from public education.1 In addition to ensuring that all children with disabilities have available a free appropriate public education and related services, Congress was concerned that distributing funds to states based on the number of students with disabilities would create incentives to over-identify. Therefore Congress stipulated: that in the reporting of the number of handicapped children being served for purposes of the formula for allocation, no State may report more than 12 percent of its total population of children aged five to 17.2 Current Law. At the time of the 1997 reauthorization of IDEA (P.L. 105-17), Congress continued to be concerned about misidentification. To address this issue, the 1997 legislation authorized a new formula, which for funds in excess of a `trigger' appropriation (approximately $4.9 billion), would base allocations on states' population and poverty.3 The Senate report explains this formula change: The Committee developed the change in formula to address the problem of over- identification of children with disabilities. When the Act was first passed in 1975, States were not providing educational services to many children with disabilities. Therefore, Congress proposed to distribute Federal funds for special education services in order to encourage and reward States for serving eligible children .... Today, the growing problem is over identifying children as disabled when they might not be truly disabled. The challenge today is not so much how to provide access to special education services but how to appropriately provide educational services to children with disabilities in order to improve educational results for such children .... This problem is most intense with minority children, especially African-American males. Over-identification of minority children, particularly in urban schools with high proportions of minority students, remains a serious and growing problem in this Nation. The problem also contributes to the referral of minority special education students to more restrictive environments.4 In addition to formula changes, the current IDEA statute has other provisions intended to prevent misidentifying children with disabilities. These include: 1 For further information on congressional intent behind P.L. 94-142, see CRS Report 95- 669, The Individuals with Disabilities Education Act: Congressional Intent, by Nancy Lee Jones. This report is currently available from the CRS website. 2 H.Rept. 94-332 to accompany H.R. 7217, p. 12. This stipulation was based on the assumption that students with disabilities accounted for about 12% of the school-age population. Note that states had to serve all children with disabilities, even if that group accounted for more than 12% of the school-age population. 3 The new formula has been in effect since FY2000. For a discussion of the formula, see CRS Report RL30810, Individuals with Disabilities Education Act: Issue Regarding "Full Funding" of Part B Grants to States, by Richard N. Apling 4 S.Rept. 105-17, p. 9. House report language is identical. CRS-3 ! Requirements related to eligibility, evaluation, and re-evaluation procedures mandating multiple, valid techniques, including local assurances that "tests and other evaluation materials used to assess a child ... are selected and administered so as not to be discriminatory on a racial or cultural basis" (§614(b)(3)(A)(i)). ! Requirements of states and the Secretary of the Interior (regarding services for Indian children with disabilities in schools operated and funded by the Secretary of the Interior) annually to collect, examine, and report to the Secretary of Education data to determine whether "significant disproportionality" is occurring with respect to race. If evidence is found, the state or the Secretary of the Interior must review and revise as necessary relevant "policies, procedures, and practices." (§618(c))5 ! Authorization of the Secretary of Education to obtain through competitive grants knowledge related to the over- or under-identification of children with disabilities (§672(a)(3)). Continuing Concerns About Misidentification. Congress and the Administration continue to be concerned about identification issues. For example, in a recent speech Secretary of Education Rodney Paige noted: In IDEA, as with ESEA, President Bush realizes that money is ineffective if it is not tied to accountability. Money alone will not improve student performance. Reform needs money, but money also needs reform. If he just wanted money, he wouldn't have bothered to ask a superintendent to be his Secretary of Education. He asked for me because he wanted reform, and reform is what he will get. We will take a systematic look at issues like discipline, finance models, disproportionate placement of minority students, over-identification, and late identification.6 With respect to late identification of students with disabilities, as discussed below, this concern has been raised most often with regard to students identified as learning disabled. 5 The Department of Education, in its final regulations for Part B grants to states, has declined to define "significant disproportionality" on the grounds that It is expected that the determination of disproportionality will involve consideration of a wide range of variables peculiar to each State including income, education, health, cultural, and other demographic characteristics in addition to race. Prescribing how the States should determine disproportionality and take corrective action would not reflect the varied circumstances existing in each State and is not consistent with discretion afforded to States under the statute (64 F.R. 12652 (March 12, 1999). 6 Remarks as prepared for delivery by U.S. Secretary of Education Rodney Paige to the Learning Disabilities Summit, Washington, D.C., August 27, 2001 (emphasis added). CRS-4 Categorizing Disabilities A discussion of identification and misidentification requires an understanding of how IDEA defines who is eligible for special education. To be eligible, a student must be identified with one or more specified disabilities (see box below) and as a result, must require special education and related services.7 1. Mental retardation 2. Hearing impairments (including deafness) 3. Speech or language impairments 4. Visual impairments (including blindness) 5. Serious emotional disturbance 6. Orthopedic impairments 7. Autism 8. Traumatic brain injury 9. Other health impairments 10. Specific learning disabilities Some disabilities have obvious physical manifestations -- such as visual or hearing impairments -- that suggest, even to an untrained individual, that a child has a disability. For example, Sherwin Nuland describes the reactions of a mother to her newborn child with Down syndrome: But as soon as I looked at him, I knew. I told myself, There's something wrong -- this child has Down syndrome.... The obstetrician said I was mistaken, but I knew better.... A few hours later, they came in and told me I was right.8 Other disabilities, such as learning disabilities, mental retardation, and emotional disturbance -- which account for about 70% of school-age children served by special education programs -- are not so easily identified. For these disabilities, identification depends on parent or teacher referral, professional judgment, criteria set by states or local educational agencies (LEAs), and often test performance. As a result, the determination of who is disabled (and thus eligible for special education) can differ from one time to another, from state to state, and possibly even from school to school. Reschly points out that It is entirely possible for students with identical characteristics to be diagnosed as disabled in one state, but not in another, or to have the categorical designation change with a move across state or school district lines.9 7 Section 602(3). At state and local discretion, for children 3-9 the definition may include those experiencing "developmental delay." In addition to these enumerated disabilities, data are often collected on those with more than one disability: deaf-blind children and those with multiple disabilities. 8 Nuland, Sherwin B. How We Live. New York, Vintage Books, 1997. p. 145. 9 Reschly, Daniel J. Identification and Assessment of Students with Disabilities. The Future of Children, v. 6, no. 1, spring 1996. p. 42. (Hereafter cited as Reschly, Identification and Assessment.) CRS-5 For example, determination of mental retardation usually is based, at least in part, on performance on an IQ test. Reschly notes that state IQ cut-off scores for mental retardation vary from 69 to 85.10 Thus a child may be judged to be mentally retarded and receive special education and related services in one state but, after moving to another state, might no longer be eligible for special education, even though the IQ score remains unchanged. State criteria may be influenced by standards set by professional associations. When professional standards change, state changes may follow, and some who were considered disabled may no longer be. For example, the American Association of Mental Retardation (AAMR) recommends an IQ cut-off score for determining mental retardation, and this score has changed over time. From 1959 to 1973, a student [scoring] one standard deviation below the mean was likely to be classified as retarded; after 1973 [when the AAMR changed the cut-off to two standard deviations below the mean] that same student was not considered retarded.... Thus, many students ceased to be retarded simply by a pen stroke of the American Association of Mental Retardation.11 Although learning disability, mental retardation, and emotional disturbance appear to be discrete disabilities, students categorized with "mild" forms12 of one of these disabilities may share many similar characteristics with those categorized as having mild forms of the other disabilities. As Ysseldyke, et al. observe, students categorized with one of these disabilities often exhibit: ! Low academic achievement, including problems with reading, mathematics, and writing, ! Delayed language acquisition, and ! Hyperactivity or inattention.13 In addition, the services provided those identified as learning disabled, mentally retarded, or emotionally disturbed might be very similar. Reschly points out instruction for these groups often utilizes the same principles and often the same procedures (intensive individualized instruction, along with close monitoring and feedback) regardless of whether the student is classified as learning disabled, mildly mentally 10 Reschly, Identification and Assessment, p. 42. 11 Ysseldyke, James E., Bob Algozzine, and Martha L. Thurlow. Critical Issues in Special Education. 3rd ed., p. 100. (Hereafter cited as Ysseldyke et al., Critical Issues.) 12 It is important to realize that "mild" disabilities are not trivial or necessarily short-lived. For example, a learning disability that inhibits the ability to read can "constitute formidable barriers to both education and occupational attainment and significantly limit adult career opportunities." Reschly, Identification and Assessment, p. 45. 13 Ysseldyke et al., Critical Issues, p. 10. CRS-6 retarded, seriously emotionally disturbed, a slow learner, or educationally disadvantaged.14 Although children with similar characteristics may be identified with different disabilities (or with no disability), the educational prospects for those in various categories may be different. Children with very similar characteristics might be classified as learning disabled in one school district, as emotionally disturbed in another district, and as "educationally disadvantaged" in a third. Each of these children might receive similar educational programs, although only the first two would be considered "disabled" under IDEA and thus covered by all the legal protections of the Act. However, teachers' and parents' perceptions of these children could be quite different, resulting in different stigmas and potentially quite different educational outcomes. The first child might be seen as having "normal" ability but needing to cope with or compensate for a learning disability. The third child might be seen as a "slow learner" or a "late bloomer" who may "catch up" in the future. But the child categorized as emotionally disturbed might face more limited educational prospects. For example, the SRI International National Longitudinal Transition Study of Students with Disabilities found that Youth classified as emotionally disturbed demonstrated a pattern of disconnectedness from school. They often were absent from school .... They had lower grade point averages than youth in other [disability] categories and were more likely to have failed courses and to have been retained in grade at the end of the year. This disconnectedness and academic failure culminated in youth with emotional disturbances having the highest dropout rate of all youth with disabilities.15 Thus the classification of a child can have significant effects. Numbers of Students Served by Special Education Data on the number of children served under IDEA and how they are classified are helpful in examining the identification and possible misidentification of children with disabilities. Each year the U.S. Department of Education (ED) collects data from states (pursuant to §618 of IDEA) on the characteristics of children receiving special education and publishes a report to Congress based on these data. The current annual report to Congress (the 22nd) provides data for school year 1998- 1999 and is the major source of data for this report.16 14 Reschly, Identification and Assessment, p. 47. 15 National Longitudinal Transitional Study of Special Education Youth. Youth with Disabilities: How Are They Doing? SRI International, Menlo Park, CA, September 1991. p. S-10. (Hereafter cited as SRI, How Are They Doing?) It is important to note that the degree of school failure among students classified as emotionally disabled may not carry over to post-school experience. The SRI study found that these students were as likely as other youth with disabilities to be employed, to earn comparable wages, and to live independently. 16 U.S. Department of Education. To Assure the Free Appropriate Public Education of All Children with Disabilities. Twenty-Second Annual Report to Congress on the (continued...) CRS-7 As background to a discussion of identifying students with disabilities, it is useful to have some sense of the flow of children into, through, and out of the special education system. In school year 1998-1999, approximately 5.5 million students ages 3-22 received special education and related services. Figure 1, which depicts the number of children served at each age in 1998-1999, shows that the number of students in special education is not evenly distributed across ages.17 In the preschool years (ages 3 and 4), relatively few children are served (between 100,000 and 200,000). The number grows in the early elementary grades, peaking about the age 10 (5th grade18) at over 500,000. Thereafter, the numbers served decline. The number of those served who are 16 is below 400,000. At age 18 fewer than 200,000 are served. At older ages those served number only a few thousand. For example, only 3,000 of the students served are 22 years of age. Figure 1. Number of Children Served Under IDEA by Age (school year 1998-1999) 600,000 500,000 Number of Disabled Students 400,000 300,000 200,000 100,000 0 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Age Source: ED, 22nd Annual Report, Table AA8. What explains the growth and decline in students served? Part of the explanation is the larger patterns of school attendance. For example, the numbers of 16 (...continued) Implementation of the Individuals with Disabilities Education Act. 2000. (Hereafter cited as the ED, 22nd Annual Report.) 17 Note that these data depict a "snapshot" of special education enrollment at some point in school year 1998-1999. While we do not have a picture of a single cohort of children as they pass through the education system, if we did have such a longitudinal perspective over 20 years, we would probably see a similar pattern of ebb and flow. 18 These data are collected and reported by age. The grade level for most children of a given age can be obtained by subtracting 5 years from their age. Thus most children age 6 are in 1st grade; most 7 year olds are in 2nd grade, etc. CRS-8 disabled children served decline by age 16, in part, because students are dropping out -- both students with disabilities and those not classified as disabled (although students with disabilities are more likely than their peers to drop out of school).19 Numbers drop further at ages 17 and 18 as students with disabilities (like their peers) graduate or complete high school. Finally some 19, 20, 21, and 22 year olds are served because of state laws providing special education for those ages.20 Numbers of Children with Various Disabilities Important trends emerge if the ebb and flow of students into and out of the special education system is examined by disability type. Figure 2 shows the numbers of students at ages 6-22 identified with the most widely recognized disabilities: specific learning disabilities, speech or language impairment, mental retardation, and emotional disturbance. Figure 2 also combines other disabilities defined in IDEA (see box on page 4) into a single category ("all other disabilities"). Several patterns are evident: ! The rapid increase and then decline of children categorized as learning disabled, ! The high initial level and then decline of those with speech and language impairments, ! The less pronounced increase and decline of the mentally retarded and emotionally disturbed categories, and ! The relatively constant number of those with other disabilities. The most notable pattern in Figure 2 is the initial relatively low number of students identified as learning disabled in 1st and 2nd grades (ages 6 and 7), the rapid growth between ages 8 and 13 (grades 3 to 8) in that category, and the decline thereafter. Initially children identified as learning disabled account for less than 50,000 special education students at age 6 (or about 12% of the overall special education population). By age 13, there are about 300,000 students identified as learning disabled or more than 60% of the total special education population. This pattern may result (at least in part) from the method often used to identify learning disabilities. Learning disabilities are often defined as a discrepancy between a child's ability and his or her achievement in areas such as reading, listening, writing, and mathematics.21 Although multiple measures are required to 19 Dropout rates can vary considerably depending on the method of calculation. Thus it is important to use comparable calculation when comparing groups. The National Longitudinal Transitional Study of Special Education Youth appears to have calculated comparable national drop rates in the mid-1980s for students with disabilities (42.9%), youth with similar demographic characteristics as students with disabilities (31.6%), and youth in the general population (24.4%). SRI, How Are They Doing?, Table 5-1, p. 5-6. 20 Part of the decline at age 18 probably results because California's mandatory age limit for serving disabled students is 18. 21 See Ysseldyke et al., Critical Issues, p. 76, and IDEA regulations at §300.541. CRS-9 determine eligibility for special education under IDEA,22 a predominant approach to determining whether a child has a learning disability is comparing an IQ test score (as a measure of "ability") with an achievement test score for reading or math. A substantial difference in scores is used to determine that a child is learning disabled. Some researchers point out that these kinds of comparisons cannot be made much before 4th grade when "meaningful" scores can be compared. Thus a child might be learning disabled or be at risk of developing a specific learning disability but not be identified and receive special education and related services until he or she is well into elementary school. The data shown in Figure 2 supports the view that learning disability identification is delayed. Some argue that alternative early identification (perhaps by kindergarten) and early intensive intervention could prevent a learning disability from developing into a condition requiring special education. Those supporting this position also argue that waiting until the child is well into his or her schooling can mean that the disability is more intractable and that the learning disability may well persist into adulthood.23 A different pattern is evident for students identified with speech or language impairment. In the early grades (1st and 2nd grades, ages 6 and 7), they number more than 200,000 and account for more than 50% of all students receiving special education and related services. By age 14 they number about 20,000 and account for about 4% of all special education students. This pattern results because most speech and language impairments are mild and tend to diminish, or disappear, as these children mature and receive speech/language therapy. In short, many children "outgrow" these disabilities and no longer require special education. Only those with more severe impairments continue to require and receive services. Patterns of growth can also be seen for those identified as mentally retarded and emotionally disturbed, although these patterns are overshadowed in Figure 2 by the pronounced swings in learning disabilities and speech/language impairment. For those age 6 (1st grade), about 22,000 are identified as mentally retarded and 9,000 are identified as emotionally disturbed. By age 13 (8th grade), the number identified as mentally retarded doubles (53,000) and the number identified as emotionally disturbed has grown to nearly 49,000. 22 IDEA Section 614(b)(2)(A) and (B) require that "a variety of assessment tools and strategies" be used to evaluate students for eligibility and prohibits the use of "any single procedure as the sole criterion for determining whether a child is a child with a disability...." 23 For an exposition of this argument, see Reid, Lyon, G., et al. Rethinking Learning Disabilities. In Chester E. Finn, Jr, et al. Rethinking Special Education for a New Century. The Thomas B. Fordham Foundation and the Progressive Policy Institute, May 2001. p. 259-307. CRS-10 Figure 2. Number of Children with Various Disabilities by Age (school year 1998-1999) 350,000 300,000 Students with Disabilities 250,000 200,000 150,000 100,000 50,000 0 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Age SPECIFIC LEARNING DISABILITIES SPEECH OR LANGUAGE IMPAIRMENTS MENTAL RETARDATION EMOTIONAL DISTURBANCE ALL OTHER DISABILITIES Source: ED, 22nd Annual Report, Table AA8. CRS-11 The pattern for those classified under other disabilities is more stable. This is particularly true if the category of "other health impairment" (which can include children with attention deficit disorder (ADD)/attention deficit hyperactivity disorder (ADHD)) is removed,24 leaving students with visual, hearing, orthopedic, and other physical impairments. These students account for about 4% of the service population at all ages until the last few years of schools. In those years (ages 19-22), these students account for a growing percentage -- nearly 13% of the 22 year olds served.25 Many of these students have relatively severe disabilities and continue to receive services until they reach the age when eligibility ends. Gender Differences Among Students Identified as Disabled Even though males and females account for roughly the same percentages of school-age population, males are more frequently identified as disabled. Figure 3 shows that males account for substantially higher percentages of the more prevalent disability groups. Male students make up about 70% of learning disabled, about 60% of mentally retarded, and about 80% of emotionally disturbed. Males also account for higher proportions of several other disability groups. ED posits several possible explanations for higher male identification: ! Physiological differences, such as higher rates of birth defects among boys and slower maturation rates; ! Teacher bias: e.g., evidence that female teachers are more likely than male teachers to refer boys for special education coupled with the predominance of female teachers in the teaching force, especially in the elementary grades; ! Assessment bias: e.g., girls may be under-represented in the category of emotional disturbance because assessment instruments are often poorly designed to detect some emotional problems, such as depression, that evidence suggests occur at higher rates among girls than among boys.26 24 As Ysseldyke et al. note (Critical Issues, p. 72): In many states, the increase [in the number of students with other health impairments] was primarily due to increased provision of services to students with attention deficit disorder .... It is also due to improved health care technology -- babies with significant impairment are living well beyond infancy. 25 Recall that the overall service population is declining because of dropping out and graduation. Thus children in these other categories account for higher percentages of a much smaller total disabled student population. 26 For further discussion and references to studies, see U.S. Department of Education. To Assure the Free Appropriate Public Education of All Children with Disabilities. Twentieth Annual Report to Congress on the Implementation of the Individuals with Disabilities Education Act. 2000, p. II-27­II-30. CRS-12 Figure 3. Gender Breakdown of Three Disability Groups (elementary and secondary students, 1994 data) 100% 90% 80% 70% 60% Female 50% 40% Male 30% 20% 10% 0% Learning Mental Emotional Disabilites Retardation Disturbance Source: U.S. Department of Education. To Assure the Free Appropriate Public Education of All Children with Disabilities. Twentieth Annual Report to Congress on the Implementation of the Individuals with Disabilities Education Act. 2000, Table II-5, p. II-27. The table presents data from the Department of Education Office Civil Rights 1994 Elementary and Secondary School Compliance Report. Distribution of Disabilities by Race/Ethnicity We turn now to a discussion of differential rates of identification for special education among racial and ethnic groups. We will examine rates from two perspectives: The rates of occurrence of various disabilities among minority students with disabilities (for example, what percentage of African Americans with disabilities are identified as mentally retarded? -- Figure 4); and the percentages of various disability groups that these groups account for (for example, what percentage of those identified as mentally retarded are African American?-- Figures 5 and 6). Disability Rates Among Racial/Ethnic Groups. Figure 4 shows the rates at which all students with disabilities ages 6-21 and disabled students of various racial and ethnic backgrounds are classified by specific disabilities.27 In general, students with learning disabilities account for the highest percentage in all groups, but the percentages differ from group to group. The learning disabled account for about 50% of all students with disabilities and for whites with disabilities; for a higher percentage of all American Indians with disabilities (about 55%); and of Hispanics with disabilities (about 60%). Learning disabled students account for somewhat lower percentages of Asian Americans/Pacific Islanders with disabilities (about 40%) and for African Americans with disabilities (about 45%). With the exception of African Americans with disabilities, speech/language impairment is the 27 Data for Hispanics in Figures 4, 5, and 6 include all races; whites and other racial/ethnic groups exclude Hispanics. CRS-13 next most prevalent category of students. For African Americans, the second most prevalent category is mental retardation. Figure 4. Composition of Population with Disabilities by Racial/Ethnic Groups (school year 1998-1999) 100% 90% 80% 70% Learning Disab. 60% Speech/Language 50% Mental Retardation 40% Emot . Disturbance 30% All Other Disabilities 20% 10% 0% All Disabled American Indian Asian/Pacific African Hispanic White St udents Islander American Source: ED, 22nd Annual Report, Table AA3. Racial/Ethnic Composition of Disability Groups. Another perspective is to examine the composition of various disability categorizations with respect to race/ethnicity. Figure 5 shows the percentages of students ages 6-21 in various disability categories who are American Indians, Asian Americans and Pacific Islanders, African Americans, and Hispanics. Because whites account for more than 60% of students with disabilities, they are excluded from this figure so that patterns for other groups can be seen more easily. (Patterns for whites in comparison to those of other groups are discussed in connection with Figure 6.) A useful standard against which to compare the classification of various racial and ethnic groups is the proportion of the overall relevant population they account for. The first set of vertical bars in Figure 5 show the percentages of the total population ages 6-21 that each of the four groups accounts for. African Americans and Hispanics each account for about 15% of this group. Asian Americans and Pacific Islanders make up about 4% of this population. American Indians account for about 1%. The remaining groups of four vertical bars show the distribution of these four racial/ethnic groups by disability, namely, "All Disabilities," "Learning Disabilities," "Speech/Language," "Mental Retardation," "Emotional Disturbance," and "All Other Disabilities." One can get some sense of the representation of each race/ethnic group by comparing the group's percentage of the total population with its percentage of a disability group. For example, African Americans account for about 15% of the total population ages 6-21 but about 20% of all students with disabilities. Thus the CRS-14 rate at which African Americans are identified as disabled is greater than their percentage of the total relevant population. On the other hand, Asian Americans/Pacific Islanders rate of identification is lower than their percentage of the general population -- accounting for about 4% of the general population and about 2% of all students with disabilities. Using the same standard, American Indians' rates are higher, and Hispanics account for about the same percentage of the overall population and for students with disabilities. The most noticeable disparities in the data presented in Figure 5 are those with respect to African Americans identified as mentally retarded and emotionally disturbed. While African Americans account for about 15% of the total population ages 6-21 and about 20% of all students with disabilities in this age group, they account for over 25% of students identified with emotional disturbance and nearly 35% of those identified as mentally retarded. CRS-15 Figure 5. Racial/Ethnic Composition of Total Population Ages 6-21, of All Students With Disabilities, and of Populations of Students with Selected Disabilities (school year 1998-1999) 40.0% 35.0% 30.0% 25.0% American Indian 20.0% Asian/Pacific Islander 15.0% African American 10.0% Hispanic 5.0% 0.0% s ge b. s e ie n tie c 1) isa a tio lit an gu ili -2 bi D a rb an rd b (6 isa ng isa u /L ta ist n D ni Re tio rD ch D ar ll ee la A l e . Le ta ot th pu Sp en Em O Po M ll A l ta To Source: ED, 22nd Annual Report, Table AF8. Note: Whites have been excluded so that patterns for other groups can be seen more easily. CRS-16 Although Figure 5 gives some perspective on possible over- and under- representation among racial and ethnic groups, a clearer picture can be obtained by dividing each group's percentage share of a disability category by the group's percentage share of the total population ages 6-21. This provides a common metric for each group and allows comparisons between groups that account for vastly different shares of the overall population (for example, comparing whites and American Indians, which is difficult without a common metric because the former accounts for nearly 70 times more of the population than does the latter). Based on this metric, the measure for a group will be 1.00 if its share of a disability category is the same as its share of the total population; less than 1.00 if its share of the category is less than its share of the total population (possible under-representation); and more than 1.00 if the category share is greater than the population share (possible over-representation). Figure 6 shows these comparisons for the four groups in Figure 5 plus whites. The first vertical bar in each group compares a group's share of all students with disabilities with the group's share of the total population ages 6-21. For American Indians, the share of all disabled students is about 1.25 times the group's share of the total population. For Asian Americans, the share of all disabled students is less than 50% of the group's share of the total population. For African Americans, the ratio is about 1.35 to 1.00. For Hispanics and whites the ratio is close to 1.00 to 1.00. That is, both these groups account for about the same shares of students with disabilities as they do of the total population ages 6-21. Once again, the most noticeable features of Figure 6 are the ratios of African Americans' shares of students identified as mentally retarded (about 2.25 times the share of the overall population) and as emotionally disturbed (about 1.75 times the share of the overall population). While comparing shares of overall population to shares of disability groups is useful, it is important to note that a ratio above 1.00 does not necessarily indicate that a group is being systematically over-identified or that a ratio below 1.00 necessarily means systematic under-identification. Because the occurrence of some disabilities are correlated with individual characteristics, such as poverty, one would expect the incidence of disabilities to be higher among groups with greater incidence of those characteristics. Thus, for example, it is not unreasonable to expect that American Indians and African Americans -- given their higher rates of poverty on the whole -- would account for somewhat higher proportions of some disability categories than of total population.28 Although it may be possible to explain some groups' relatively high rates of disabilities based on poverty and other factors related to disabilities, it is difficult to conclude that the high rates of African Americans categorized as mentally retarded 28 Based on data from the March 2000 Current Population Survey (CPS) conducted by the Bureau of the Census, African Americans account for about one-third of poor children below the age of 18. The statistical relationship between low income and disability is clear. For example, the SRI National Longitudinal Study found that about 35% of youth with disabilities had family incomes below $12,000; while about 18% of all youth came from families with incomes below that level (based on 1987 data). (SRI, How Are They Doing?, Table 2-16, p. 2-26). CRS-17 and emotionally disturbed can be completely explained by such factors alone. For one thing, if general disadvantage were the main explanation, one would expect the rates of identification for a racial or ethnic group to be similar across various disabilities. However, African Americans, for example, account for nearly twice the percentage of students with mental retardation as they do of students with learning disabilities (34% vs. 18%). In addition, one might expect that identification rates would be similar for all disadvantaged groups. However, for Hispanic students, who tend to have poverty rates similar to African Americans, the share of students classified as mentally retarded is less than their overall share of the student population. Moreover, it is important to remember that these data are for aggregate groups. Even if whites, for example, represent about the same percentage of students with disabilities as they do of the overall student population, this does not necessarily mean that no white student is misidentified as disabled or is misclassified regarding the type of disability. While aggregate state data suggest some areas of misidentification of students with disabilities and may mask other problem areas, it would appear that the areas of most concern are the unusually high rates of identification of African Americans as mentally retarded and as emotionally disturbed and the comparatively low rates of identification of Asian Americans and Pacific Islanders as students with disabilities. While there is no complete explanation for such outcomes, many maintain that substantial numbers of African Americans are being incorrectly identified as mentally retarded and as emotionally disturbed. The explanation for some misidentification may be selection procedures. In the case of mental retardation, the use of IQ tests, which may exhibit subtle bias against some minority groups, as the predominant tools for identifying students as mentally retarded could be a contributing factor.29 The subjective judgment of teachers could be another factor. Ysseldyke and his colleagues have done extensive research on the process by which students are referred to and evaluated for special education. They have found that it is likely a student will be deemed eligible once a classroom teacher or parent has referred a child for evaluation. They are disturbed by the direct link between referral and eligibility because referrals often are unrelated to students' actual behavior and abilities. We have demonstrated repeatedly that teachers refer students who bother them. Different teachers are bothered by different kinds of behavior, so the process is very unpredictable.30 To the extent that eligibility for special education follows from teacher referral that may be based on preconceived notions, "bothersome" deportment, or prejudice rather 29 See Ysseldyke et al., Critical Issues, p. 74. Noted here, very low scores on an IQ test should not be the only justification for identifying a student as mentally retarded. Adaptive behavior, i.e, how well a student functions in the social environment should also be considered. However, judging adaptive behavior is subjective; whereas an IQ score is an "objective" measure. 30 Ysseldyke, Jim. Reflections on a Research Career: Generalizations from 25 Years of Research on Assessment and Instructional Decision Making. Exceptional Children, v. 67, no. 3, spring 2001. p. 303. CRS-18 than on objective assessments of students' abilities and behavior, one can imagine how misidentification could occur. If a student "bothers" a teacher and classification as mentally retarded or emotionally disturbed means the student will be placed outside the regular classroom, one can imagine that referral for eligibility assessment could be seen as an easy remedy for removing a "trouble maker."31 There is less extensive explanation for the low rates of identification for Asian Americans/Pacific Islanders. Is the actual occurrence of disabilities significantly lower among this group than in the overall population? Or are significant numbers of Asian Americans/Pacific Islanders being under-identified as requiring special education and related services and thus being denied services guaranteed by IDEA? 31 Arguably a contributing factor is the mismatch between growing minority populations in many public school districts and teaching forces in most districts that are predominately white. For example, data show that students from minority backgrounds account for nearly 70% of total enrollment in low-income urban school districts, while the teaching force in those districts is about 70% white. Clewell, B. C., and A. M. Villegas. Introduction. Special Issue on Diversifying the Teaching Force to Improve Urban Schools: Meeting the Challenges. Education and Urban Society, v. 31, no. 1, November 1998. p. 4. CRS-19 Figure 6. Ratio of Percentage of A Disability Group to Percentage of Total Population for Racial/Ethnic Groups (school year 1998-1999) 2.50 2.00 All Disabilities 1.50 Learning Disab. Speech/Language 1.00 Mental Retardation 0.50 Emot. Disturbance All Other Disabilities - American Asian/Pacific African Hispanic White Indian Islander American Race /Ethnicity Source: ED, 22nd Annual Report, Table AF8. CRS-20 Possible Policy Alternatives Congress has recognized problems of misidentification and has provided many remedies in the original Act and added to those remedies in the current IDEA. In general, an overarching aim of IDEA has been and is to ensure that children with disabilities are properly identified and provided with free appropriate public education. The extensive statutory and regulatory provisions related to "child find," evaluation and re-evaluation, and the Individualized Education Program process are all intended and designed to fulfill this aim. Given that Congress approved extensive IDEA amendments relatively recently (June 1997) and final regulations have been in effect for less than 3 years (March 1999), some argue that further amendments to IDEA to address issues such as misidentification of children with disabilities should be postponed until current law and regulation can be fully implemented and evaluated.32 Related to the view of postponing further changes in IDEA is the possibility of improved enforcement of current law. Thomas Hehir at a recent hearing on over- representation recommended improved monitoring and enforcement of IDEA as one remedy: Though there are powerful monitoring and enforcement mechanisms in IDEA, they are not self implementing. They require an active federal presence. The history of federal enforcement of IDEA has been relatively weak. One reason for this is the relative small size of the federal work force devoted to this issue, about fifty people. Though I believe a significantly larger workforce is justified in this area, the more important factor is the willingness of both administrations [sic] and Congress to support strong federal enforcement. When I was at OSEP [Office of Special Education Programs at ED], members of Congress from the states involved in enforcement actions immediately opposed several enforcement actions, brought by the Department of Education. Given the fact that IDEA now addresses the issue of overrepresentation, this committee should seek a significantly enhanced federal monitoring and enforcement role and support the Department in doing its job.33 Others maintain that IDEA requires further consideration to address misidentification and other issues.34 While a comprehensive review of possible legislative action is beyond the scope of this report, some possibilities include: 32 This was an argument made by Thomas Hehir (formerly Director of the Office of Special Education Programs at the Department of Education) at a hearing before the House Education and the Workforce Committee, October 4, 2001. 33 Ibid., pp. 3-4. For an overview of how current law addresses misidentification, see the discussion of current law earlier in this report. 34 For example, a fact sheet from the House Education and the Workforce Committee dated September 10, 2001 states that "Congress will turn its attention to reauthorization and reform of IDEA" once the reauthorization of the Elementary and Secondary Education Act is completed. CRS-21 ! Greater emphasis on, together with rigorous evaluation of, early intervention -- including possible increased funding for the preschool (Part B, Section 619) and infant and families grants to states (Part C), which have received few increases in recent years. As some have argued, earlier assessment and identification of disabilities coupled with earlier intervention could reduce the prevalence of some disabilities, especially learning disabilities. ! Additional technical assistance and remedial action from the federal government to the states and from states to school districts based on evidence of misidentification under §618(c) of IDEA -- for example, requiring states to develop and implement "corrective action plans" that would be submitted to and approved by the Secretary, as suggested by one commenter on the IDEA regulations.35 ! Additional requirements for teacher pre-service and in-service training of teachers, especially the training of regular classroom teachers in proper identification of children with disabilities for referral for evaluation and with respect to better strategies for accommodating children with disabilities in the regular classroom so that regular classroom teachers are more confident in serving children with disabilities and less prone to refer them for evaluation in the hopes of removing these children from regular classrooms. ! Further research regarding, for example, the possible under-representation of some groups, such as Asian Americans and Pacific Islanders. 35 64 F.R. 12652 (March 12, 1999). ------------------------------------------------------------------------------ For other versions of this document, see http://wikileaks.org/wiki/CRS-RL31189