Social Security: New Functional Assessments for Children Raise
Eligibility Questions (Letter Report, 03/10/95, GAO/HEHS-95-66).

Pursuant to a congressional request, GAO reviewed the effects of the
judicially mandated indivualized functional assessment (IFA) process on
Supplemental Security Income (SSI) benefits, focusing on: (1)
allegations that parents may be coaching their children to fake mental
impairments to qualify under the lower eligibility standards created by
IFA; and (2) how IFA affects the children's eligibility for benefits.

GAO found that: (1) the judicial decision that required changes in IFA
essentially made the process for determining disability in children
analogous to the adult process; (2) the new process assesses how
children's impairments limit their ability to act and behave like
unimpaired children of similar age; (3) it has become important to
obtain evidence of disability from nonmedical sources as part of the
children's assessment; (4) although the court required a new type of
assessment for disabled children, it did not define the degree of
limitation necessary to qualify for SSI benefits; (5) before the IFA
process was introduced in 1991, the national award rate for all types of
childhood cases was 38 percent, but the award rate jumped to 56 percent
in the first 2 years after IFA regulations were issued; (6) the
non-medical aspects of the IFA evaluation relies heavily on adjudicator
judgment; (7) while the Social Security Administration (SSA) has
attempted to improve the process, and thereby reduce fraud and improve
accuracy in awards, IFA has an underlying conceptual problem; (8)
although the IFA process attempts to improve accuracy, the presence of
coaching by parents is almost impossible to detect; and (9) more
consistent eligibility decisions could be made if adjudicators based
functional assessments of children on the functional criteria in SSA
medical listings.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-95-66
     TITLE:  Social Security: New Functional Assessments for Children 
             Raise Eligibility Questions
      DATE:  03/10/95
   SUBJECT:  Handicapped children
             Fraud
             Social security benefits
             Children
             Eligibility criteria
             Eligibility determinations
             Disability benefits
             Mental illnesses
             Judicial opinions
IDENTIFIER:  Food Stamp Program
             Supplemental Security Income Program
             Medicaid Program
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

March 1995

SOCIAL SECURITY - NEW FUNCTIONAL
ASSESSMENTS FOR CHILDREN RAISE
ELIGIBILITY QUESTIONS

GAO/HEHS-95-66

SSI:  Childhood Eligibility Decisions


Abbreviations
=============================================================== ABBREV

  DBRA - Disability Benefits Reform Act of 1984
  DDS - disability determination service
  HHS - Health and Human Services
  IFA - individualized functional assessment
  IG - Inspector General
  IQ - intelligence quotient
  RFC - residual functional capacity
  SSA - Social Security Administration
  SSI - Supplemental Security Income

Letter
=============================================================== LETTER


B-257473

March 10, 1995

The Honorable Herb Kohl
The Honorable David Pryor
United States Senate

The Honorable George W.  Gekas
The Honorable Gerald D.  Kleczka
The Honorable Blanche Lambert Lincoln
The Honorable Nick Smith
House of Representatives

The number of children receiving Supplemental Security Income (SSI)
benefits has nearly tripled over the last 5 years from 300,000 to
almost 900,000, and benefit payments now exceed $4 billion annually. 
The Social Security Administration (SSA) awards SSI benefits to
disabled children who live in families with low incomes and limited
resources.  A number of factors have contributed to the growth in
children's awards, including outreach efforts by SSA and child
advocates, rising numbers of children in poverty, and major changes
in the criteria for determining whether children are disabled. 
Growth has been especially rapid in awards to children with mental
impairments. 

Particularly troublesome have been allegations that parents coach
their children to fake mental impairments by misbehaving or doing
poorly in school so that they can qualify for cash benefits.  These
benefits can amount to almost $5,500 per year for each disabled
child.\1 The coaching allegations, which have been widely reported by
the media, have created the perception among the public that the
program is vulnerable to fraud and abuse.  In addition, concerns have
been raised that the program could foster lifelong dependence on
government assistance if children come to view the label "disabled"
as a lifetime entitlement to income and medical benefits.  Finally,
concerns have been raised about whether the program's eligibility
criteria for children are too lenient.  As a result of these
concerns, reform of the SSI childhood disability program is now the
subject of congressional scrutiny. 

In our October 21, 1994, briefing, you asked us to report on SSA's
new way of assessing children's impairments using the individualized
functional assessment (IFA) process mandated by the Supreme Court in
Sullivan v.  Zebley.  The new IFA process permits the award of
benefits to children with impairments that are less severe than the
impairments that previously could justify an award.  We assessed (1)
the IFA's impact on the SSI rolls, (2) its implementation by SSA, and
(3) its vulnerability to coaching. 

To develop the information in this report, we (1) reviewed SSA's
childhood disability program policies, procedures, and records, and
discussed the IFA process with SSA program officials on the national,
regional, and local level; (2) interviewed officials in state
disability determination services (DDSs); (3) reviewed SSA's study of
decisions made on childhood cases involving behavioral and learning
disorders; and (4) attended a June 1994 SSA training course designed
to address the problems raised in the study.  We also discussed
eligibility issues with officials of the Department of Health and
Human Services' (HHS) Office of Inspector General (IG), which
recently issued two reports on the SSI childhood disability program. 
(See app.  I for more details on our scope and methodology.)


--------------------
\1 Benefits generally are provided without regard to the number of
SSI recipients in the household.  SSA estimated in March 1994 that
125,000 children receiving SSI lived in households with at least one
other SSI recipient. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

Changes in the regulations governing childhood eligibility for SSI
have had a significant impact on the growth and composition of the
childhood disability rolls.  In particular, awards have been made to
more than 200,000 children who did not meet SSA's listing of
impairments but instead qualified for benefits based on the less
restrictive IFA criteria.  These awards account for about $1 billion
a year in benefit payments.  About 84 percent of the children
qualifying based on IFAs have mental impairments, and about one-half
of the awards for behavioral disorders, including attention deficit
hyperactivity disorder, are based on the IFA criteria. 

In our analysis, we found fundamental flaws in the IFA process. 
Specifically, each step of the process relies heavily on
adjudicators' judgments, rather than objective criteria from SSA, to
assess the age-appropriateness of children's behavior.  As a result,
the subjectivity of the process calls into question SSA's ability to
ensure reasonable consistency in administering the SSI program,
particularly for children with behavioral and learning disorders. 

In addition, rapid program growth, particularly in the award of
benefits to less severely impaired children, may also have
contributed to the public concern that parents could be coaching
their children to fake mental impairments in order to qualify for
benefits.  Studies that we reviewed, however, have found little
evidence that coaching is widespread, but they relied solely on
documentation in case files and, therefore, cannot rule out coaching. 
Although coaching allegations are troublesome, substantiating them
and measuring the extent of coaching is virtually impossible. 


   BACKGROUND
------------------------------------------------------------ Letter :2

Since 1974, the SSI program, under title XVI of the Social Security
Act, has provided benefits to low-income blind and disabled
persons--adults and children--who meet financial eligibility
requirements and SSA's definition of disability.  SSA determines
applicants' financial eligibility; state DDSs determine their medical
eligibility.  DDSs are state agencies that are funded and overseen by
SSA.  To meet the financial test, children must be in families with
limited incomes and assets. 

In 1994, children's federally administered SSI payments totaled $4.52
billion.  Depending on the family's income, an eligible child can
receive up to $458 per month in federal benefits; 27 states also
offer a supplemental benefit payment.  Because SSI is an individual
entitlement, no family cap exists on the amount of benefits received
in a household.  With SSI eligibility usually come other in-kind
benefits, most notably Medicaid and Food Stamps. 

The Social Security Act defines a disabled child as a person under
age 18 who "suffers from any medically determinable physical or
mental impairment of comparable severity" to one that disables an
adult.  The statute defines adult disability in terms of an inability
to work either in a former job or in any other job in the national
economy.  Specifically, adult disability is defined as the inability

     "to engage in any substantial gainful activity by reason of any
     medically determinable physical or mental impairment which can
     be expected to last a continuous period of not less than twelve
     months."

Because children are not expected to work, however, this definition
is not applicable to measure disability in children. 

At a DDS, childhood disability determinations are made by an
adjudication team consisting of an examiner and a medical consultant. 
For mental impairments, the consultant must be a psychiatrist or
child psychologist.  The examiner collects all medical
evidence--physical and mental--either from medical sources who have
treated the applicant or from an independent consultant if more
medical information is needed.  The examiner supplements the medical
information with accounts of the child's behavior and activities from
the child's teachers, parents, and others knowledgeable about the
child's day-to-day functioning. 

Working together, the DDS adjudication team determines whether the
applicant's medical condition matches or is equivalent to an
impairment found in SSA's listing of medical impairments.\2 If so,
benefits are awarded.  If, however, the applicant's condition is not
severe enough to meet or equal the severity criteria in SSA's medical
listings, the team uses the evidence to perform an IFA.  If the IFA
shows the child's impairment substantially reduces his or her ability
to function age-appropriately, benefits are awarded.  If not, a
denial notice is issued, and applicants are informed of their appeal
rights. 


--------------------
\2 SSA's listing of medical impairments describes impairments--in
terms of signs, symptoms, and laboratory findings--that are presumed
to be severe enough to disable an individual. 


   SSI CHILDHOOD ELIGIBILITY
   CRITERIA HAVE UNDERGONE MAJOR
   CHANGES
------------------------------------------------------------ Letter :3

During a 2-month period, SSA issued two sets of new regulations that
significantly changed the criteria for determining children's
eligibility for SSI disability benefits.\3 One set of regulations,
issued in accordance with the Disability Benefits Reform Act of 1984
(DBRA), revised and expanded SSA's medical listings for evaluating
mental impairments in children to incorporate recent advances in
medicine and science.  The second set of regulations was issued in
response to the Sullivan v.  Zebley Supreme Court decision, which
required SSA to make its process for determining disability in
children analogous to the adult process.  Both sets of regulations
placed more emphasis on assessing how children's impairments limit
their ability to act and behave like unimpaired children of similar
age.  Both also emphasize the importance of obtaining evidence from
nonmedical sources as part of this assessment. 


--------------------
\3 For a complete description of these changes, see Social Security: 
Rapid Rise in Children on SSI Disability Rolls Follows New
Regulations (GAO/HEHS-94-225, Sept.  9, 1994). 


      DBRA REGULATIONS CHANGED
      SSA'S MEDICAL LISTINGS FOR
      ASSESSING MENTAL IMPAIRMENTS
      IN CHILDREN
---------------------------------------------------------- Letter :3.1

SSA issued new regulations in accordance with DBRA on December 12,
1990.  These new regulations revised and expanded SSA's medical
listings for childhood mental impairments to reflect up-to-date
terminology used by mental health professionals and recent advances
in the knowledge, treatment, and methods of evaluating mental
disorders in children.  The new medical listings for mental
impairments provided much more detailed and specific guidance on how
to evaluate mental disorders in children than the former regulations,
which were published in 1977.  In particular, the new medical
listings placed much more emphasis on assessing how a child's mental
impairment limits his or her ability to function in age-appropriate
ways.  SSA made this change because mental health professionals
consider functional factors particularly important in evaluating the
mental disorders of children. 

The former medical listings for mental impairments emphasized the
medical characteristics that must be met to substantiate the
existence of the impairment.  Specific areas of functioning sometimes
were and sometimes were not mentioned as a factor in this
determination.  In contrast, the new medical listings provide much
more detailed guidance on assessing the functional aspects of each
impairment.  The standard for most impairments is divided into two
parts:  medical and functional criteria, both of which must be
satisfied for the applicant to qualify for a benefit. 

The functional criteria are described in terms of the age of the
child and the specific areas of functioning--such as social,
communication/
cognition, or personal/behavioral skills--that must be assessed.  The
new medical listings emphasize the importance of parents and others
as sources of nonmedical information about a child's day-to-day
functioning.  In general, the childhood mental listings require
children over 2 years old to have marked limitations in two of the
four areas of functioning to qualify for benefits.  Further, when
standardized tests are available, the listing defines the term
"marked" as a level of functioning that is two standard deviations
below the mean for children of similar age. 

The new medical listings also classified childhood mental disorders
into more distinct categories of mental impairments.  Previously,
there were 4 impairments listed--mental retardation, chronic brain
syndrome, psychosis of infancy and childhood, and functional
nonpsychotic disorders--now there are 11.  Several of the newly
listed impairments, such as autism and other pervasive developmental
disorders, mood disorders, and personality disorders, describe
impairments that were previously evaluated under one or more of the
four broader categories of childhood mental impairments.  Several
other impairments are mentioned for the first time, such as attention
deficit hyperactivity disorder and psychoactive substance dependence
disorders. 


      ZEBLEY REGULATIONS ADDED
      SEPARATE FUNCTIONAL
      ASSESSMENT PROCESS
---------------------------------------------------------- Letter :3.2

On February 20, 1990, the Supreme Court ruled that SSA's process for
determining disability in children under 18 years old violated the
Social Security Act because the process held children to a more
restrictive disability standard than it did adults.  In its opinion,
the Court found that the process for children

     "does not account for all impairments 'of comparable severity'
     [to adults], and denies
     child claimants the individualized functional assessment that
     the statutory standard
     requires .  .  .  ."

To determine adults' eligibility for disability benefits, SSA uses a
five-step sequential evaluation process.  Before Zebley, it used only
a two-step process to determine children's eligibility for benefits. 
(See fig.  1.) Children were awarded benefits only if their
impairments met or equaled the severity criteria in SSA's medical
listings.  All other children were denied benefits.  In contrast,
adults whose conditions were not severe enough to qualify under the
medical listings could still be found eligible for benefits if an
assessment of their residual functional capacity (RFC) showed that
they could not engage in substantial work.  No analogous assessment
of functioning was done for children who did not qualify under the
medical listings. 

   Figure 1:  Disability
   Evaluation Process for Adults
   Versus Children

   (See figure in printed
   edition.)

To eliminate this disparity, the Court mandated that for those
children who do not qualify for benefits under the more restrictive
medical listings, SSA must add a less restrictive individualized
assessment of how the child's impairment affects his or her ability
to function in age-appropriate ways--that is, to act or behave in
ways that children of similar ages normally do--before it could
decide whether the child was eligible for benefits.  The Court said
that although a vocational analysis does not apply to children
because they are not expected to work, SSA could make

     "an inquiry into the impact of an impairment on the normal daily
     activities of a child of the claimant's age--speaking, walking,
     dressing and feeding oneself, going to school, playing, etc."

Although the Court required the functional assessment, it did not
define the degree of limitation necessary to qualify for benefits,
except by analogy to the adult definition of disability. 

To implement the Zebley decision, SSA convened a group of experts in
April 1990 to help formulate new regulations using age-appropriate
functional criteria.  Included were experts in general and
developmental pediatrics, child psychology, learning disorders, and
early and adolescent childhood education as well as advocates from
groups such as Community Legal Services in Philadelphia (plaintiff's
counsel in the Zebley case), the Association for Retarded Citizens,
and the Mental Health Law Project.  SSA also consulted with its
regional offices and the state DDSs. 

Building on the functional criteria added to the listings after DBRA,
SSA issued regulations implementing the Supreme Court's decision on
February 11, 1991.\4 According to these regulations, for the child to
be eligible for disability benefits, the IFA must show that the
child's impairment or combination of impairments limits his or her
ability "to function independently, appropriately, and effectively in
an age-appropriate manner." Specifically, the impairment must
substantially reduce the child's ability to grow, develop, or mature
physically, mentally, or emotionally to the extent that it limits his
or her ability to (1) attain age-appropriate developmental
milestones; (2) attain age-appropriate daily activities at home,
school, play, or work; or (3) acquire the skills needed to assume
adult roles.  Although SSA officials describe these as
state-of-the-art criteria for assessing children's functioning, they
concede that many of these concepts are not clear cut. 

As a result of these regulations, DDSs now perform IFAs to assess the
child's social, communication, cognitive, personal and behavioral,
and motor skills, as well as his or her responsiveness to stimuli and
ability to concentrate, persist at tasks at hand, and keep pace.\5

Like the DBRA regulations, the IFA process requires DDSs to
supplement medical information with information about the child's
behavior and activities from the child's teachers, parents, and
others knowledgeable about the child's day-to-day functioning in
order to make these assessments.  Generally, if the IFA shows that a
child has a moderate limitation in three areas of functioning or a
marked limitation in one area and a moderate limitation in another,
benefits are awarded.  In contrast, the more restrictive functional
criteria under SSA's mental listings require two marked limitations. 

In addition to measuring functioning as part of the IFA process, the
Zebley regulations added the concept of functional equivalence to
SSA's medical listings.  Before Zebley, a child qualified for
benefits only if his or her impairment met or was medically
equivalent to the severity criteria in the listings.  After Zebley, a
child could qualify if his or her impairment was functionally
equivalent to an impairment in the medical listings, as long as there
was a direct, medically determinable cause of the functional
limitations.  The regulations provide 15 examples of conditions--such
as the need for a major organ transplant--presumed to be functionally
equivalent to the listed impairments. 


--------------------
\4 Final regulations incorporating voluminous public comments were
issued on September 9, 1993.  These regulations, which were not
substantially different from the February 1991 interim final
regulations, have a September 9, 1997, sunset date, after which time
they will no longer be effective, unless the Secretary of HHS
extends, revises, or reissues them. 

\5 Social, communication, cognitive, and motor skills are assessed
for children of all ages; personal and behavioral skills are assessed
for children 1 year old and older.  The ability to concentrate,
persist at tasks at hand, and keep pace are assessed for children 3
years old and older; responsiveness to stimuli is assessed in
children under 1 year old. 


   IFA PROCESS HAS HAD A MAJOR
   IMPACT ON THE ROLLS
------------------------------------------------------------ Letter :4

Of the 646,000 children added to the SSI rolls from February 1991
through September 1994, about 219,000 (one-third) were awarded
benefits based on the less restrictive IFA process.  If all 219,000
children receive the maximum benefit, their SSI benefits would cost
about $1 billion a year.  About 84 percent of these children had a
mental impairment as their primary limitation, and about 16 percent
had physical impairments.  (Fig.  2 shows a breakdown of the
impairments.)

   Figure 2:  Most IFA Awards Go
   to Children With Mental
   Impairments

   (See figure in printed
   edition.)

Source:  Analysis of SSA's 831 file. 

Figure 3 shows the substantial increase in the number of awards. 
Much of this increase was due to the implementation of new medical
listings for mental impairments.  The IFA process also added to the
growth in the rolls and accounted for a substantial portion of new
awards.  Figure 3 also shows that the average monthly number of
applications jumped dramatically after Zebley and has continued to
grow.  Many observers attribute this increase in applications to the
publicity surrounding Zebley, as well as to increased outreach by
SSA, some of which was congressionally mandated. 

Also, some of the increase in awards may have been attributable to
the close scrutiny of the IFA process by courts and disabled child
advocates, which some believe may have resulted in some DDSs feeling
pressured to increase their award rates during the 1991-1992 period. 
(App.  II provides a chronology of their actions.) Before the IFA
process was introduced in 1991, the national award rate for all types
of childhood cases was 38 percent, but the award rate jumped to 56
percent in the first 2 years after the IFA and DBRA regulations were
issued.  More recently, during 1993 and 1994, the award rate has
dropped dramatically.  The national award rate for 1994 was 32
percent--lower than it was in the 2 years before Zebley. 

   Figure 3:  IFA and Changes in
   Medical Listings Both
   Contribute to Growth in the
   Rolls

   (See figure in printed
   edition.)

Source:  Analysis of SSA's 831 file. 


   IFA PROCESS HAS BEEN DIFFICULT
   TO IMPLEMENT CONSISTENTLY
------------------------------------------------------------ Letter :5

Our review indicates that the IFA process has been difficult to
implement consistently and reliably, particularly for children with
mental impairments, because the process requires adjudicators to make
a series of judgment calls in a complex matrix of assessments about
age-appropriateness of behavior.  SSA and IG studies of children with
mental impairments have borne out these difficulties.  Although SSA
has tried to add rigor to the IFA process through guidance and
training, we believe that problems will likely continue because of
the difficulties inherent in using age-appropriate behavior as an
analog for the adult vocational assessment of residual functional
capacity. 


      IFA--A COMPLEX PROCESS THAT
      RELIES HEAVILY ON
      ADJUDICATOR JUDGMENT
---------------------------------------------------------- Letter :5.1

Determining disability for children with impairments that are not
severe enough to match a listed impairment can be a highly subjective
process.  SSA designed the IFA process to provide DDS adjudicators
with a structure to help them make uniform and rational disability
determinations for children with less severe impairments.  Even so,
the necessity to assess a child's ability to function
age-appropriately requires DDS adjudicators to make a series of
judgments, which we believe raises questions about the consistency
and reliability of DDS decisions.  SSA and IG studies and our
analysis document problems throughout the IFA process, especially for
mental impairments.  (See app.  III for a more detailed discussion of
the problems that SSA and the IG identified.)

Extensive evidence needed:  To make disability determinations, DDSs
use information from both medical and nonmedical sources, including
teachers, day care providers, parents, and others knowledgeable about
the child's day-to-day behavior and activities.  For the functional
assessment, observations are needed about the child's behavior over a
long period of time, so evidence-gathering can be a considerable
task.  SSA found in its 1994 study that the lack of sufficient
supporting documentation was the most common problem in its sample of
childhood disability decisions. 

School officials in particular are an important source of nonmedical
data on children's behavior over time.  Each DDS develops its own
questionnaires for eliciting the data, and inquiries are made on
virtually every applicant because this information is also used to
assess functioning under the medical listings.  We estimate that the
process now results in about 500,000 inquiries to schools each year,
a substantial reporting burden.  Some parties believe that the
open-ended questionnaire design in many states and the burden on
school officials faced with many inquiries may be contributing to
poor quality data from this key source. 

Difficulty classifying limitations:  If an IFA is needed, a
disability adjudicator must classify the child's limitations in the
appropriate areas of functioning, as shown in figure 4.  This is a
complex judgment because some areas are closely interrelated and
impairments may or may not affect functioning in more than one area. 
If, for example, evidence indicates that a child gets in fights at
school, the adjudicator must determine whether the specific behavior
is evidence of a limitation in social skills, personal and behavioral
skills, or some combination of these.  SSA found that in cases of
incorrect awards a common mistake that adjudicators made was to count
the effect of an impairment in two areas when only one was
appropriate.  This resulted in the impairment seeming more severe
than it actually was. 

Problems defining degrees of limitation:  Once the areas have been
identified, the adjudicator must judge the degree of limitation. 
Because only certain conditions--such as low intelligence quotient
(IQ)--can be objectively tested and determined, SSA has defined the
severity of limitations by comparison with expected behavior for the
child's chronological age.  Figure 4 shows the degrees of limitation
adjudicators use to assess children 3 through 15 years old.  SSA's
guidance defines a limitation in the moderate category as more than a
mild or minimal limitation but less than a marked limitation.  The
terms "mild" and "minimal" are not defined, but SSA guidance
describes an impairment in the marked category as one that
"seriously" interferes with a child's ability to function
age-appropriately, while a moderate limitation creates "considerable"
interference.  Within each category, adjudicators are expected to be
able to differentiate the degree of limitation.  For example, a
moderate rating can range from a "weak moderate" (just above a
less-than-moderate) up to a "strong moderate" (just below a marked
limitation). 

   Figure 4:  Structure of the IFA
   Process

   (See figure in printed
   edition.)

Limited guidance for summing the result:  Because the IFA process is
inherently subjective, SSA cannot provide an objective procedure for
summarizing the IFA results.  Therefore, SSA instructs adjudicators
to step back and assess whether the child meets the overall
definition of disability.  As an example to guide adjudicators, SSA
has said that an award may generally be granted if a child has a
moderate limitation in three areas.  However, SSA officials stress
that this statement assumes "three good, solid moderates," and they
characterize it as a general guideline, not a firm rule.  Also, they
stress that other possible combinations of ratings, such as two
strong moderates, could justify finding a child disabled, depending
on the individual child's circumstances.  In the end, officials
stress that adjudicators are expected to award or deny benefits based
on an overall judgment, not on any specific sum of severity ratings. 


      SSA AND IG STUDIES HIGHLIGHT
      IFA DIFFICULTIES
---------------------------------------------------------- Letter :5.2

SSA's 1994 study of 325 childhood awards highlighted the difficulties
in using the IFA process to reliably identify disabled children,
particularly children with behavioral and learning disorders.\6 In
the study, SSA's Office of Disability selected cases of 325 children
with behavioral and learning disorders who had been found eligible. 
The majority were found eligible based on IFAs.  These cases had been
decided by DDS adjudicators, based on their understanding of existing
guidance from SSA.  Then, SSA's regional quality assurance staff had
reviewed the decisions and found them accurate.  The study involved a
third group of experts in the Office of Disability who reviewed the
same cases and found inaccuracies in the decisions.  Based on their
findings, we concluded that about 13 percent of the awards reviewed
by SSA had been made to children who were not impaired enough to
qualify.  Also, another 23 percent of the awards had been made
without sufficient supporting documentation.\7

A January 1995 IG report focused on IFA-based awards to children with
mental impairments.  IG staff, with assistance from the Office of
Disability, reviewed 129 IFA-based awards for mental retardation,
attention deficit hyperactivity disorder, and other behavioral or
learning disorders.  The IG found that 17 (13 percent) of the awards
should have been denials and another 38 (29 percent) had been based
on insufficient evidence.  The IG attributed this to DDS
adjudicators' difficulty interpreting and complying with SSA's IFA
guidelines for assessing the severity of children's mental
impairments.  Many adjudicators reported that they found the SSA
guidelines unclear and not sufficiently objective.  The IG stated
that this group of children had less severe impairments than those
children determined disabled based on the medical listings, making
the assessment of their impairments' effect on their ability to
function age-appropriately more difficult. 

We observed firsthand the difficulty that adjudicators face in making
the judgments required by the IFA process for children who have
behavioral and learning disorders.  In June 1994, we attended 1-day
training sessions for DDS adjudicators and SSA's regional quality
assurance staff from across the nation.  The Office of Disability
presented the findings from its 1994 study and discussed the policies
and procedures that DDS and quality assurance staff had misapplied. 
In this training, Office of Disability staff presented case studies
of children included in the 1994 study.  After those in attendance
reviewed the evidence for each child's case, they were asked to
assess the degree to which the child's impairment limited his or her
functioning.  The attendees' opinions were tallied and in all cases
they were split.  During discussions of each case, attendees often
voiced differing views on why they believed, for example, that the
child's limitation was less than moderate or moderate, or whether a
moderate limitation was a good, solid moderate, or a weak moderate. 
In some cases, the opinion of the majority of attendees turned out to
be different from the conclusion of the Office of Disability. 

In addition to the national training in June 1994, SSA took other
steps to correct implementation problems, including (1) issuing
numerous instructional clarifications and reminders, (2) requiring
DDSs to specially code certain types of mental impairments and all
decisions based on three moderate limitations (to facilitate
selecting samples of cases for further study), and (3) establishing
more rigorous requirements for documenting awards that are based on
three moderate limitations.  The Office of Disability plans to do a
follow-up study to assess the effectiveness of its remedial efforts. 

Some experts believe that further steps could be taken to improve the
IFA process.  For example, experts we contacted commented on the need
for more complete longitudinal evaluations by professionals.  They
pointed out that more complete examinations--sometimes including
multiple visits and observations of both parents and children--would
help to address concerns about the adequacy of information from
schools and medical sources and provide higher assurance of good
decisions.  They stated that because professionals are trained to
identify malingering in mental examinations, the expanded
examinations might also help relieve concerns about coaching.  They
agreed that such examinations would raise the program's
administrative costs considerably, but because a child can receive
almost $5,500 a year in benefits (which can continue for life) they
believed that the costs would be justified. 

SSA's efforts and experts' suggestions are geared toward improving
the process rather than addressing the underlying conceptual problems
with the IFA.  The difficulties so far in implementing the IFA bring
into question whether these types of incremental actions can ensure
consistently accurate decisions for children with mental impairments,
especially behavioral and learning disorders. 


--------------------
\6 The study's sampling methodology does not permit the results to be
projected to the universe of childhood cases or to any subset of the
universe. 

\7 See appendix III for details on the study and how we calculated
these percentages. 


   EXTENT OF COACHING UNKNOWN
------------------------------------------------------------ Letter :6

The rapid growth in awards to children with mental
impairments--particularly behavioral and learning disorders--has
contributed to the public perception that the SSI program for
children is vulnerable to fraud and abuse.  The media have reported
allegations that parents coach their children to fake mental
impairments by misbehaving or performing poorly in school so that
they can qualify for SSI benefits.  Critics believe that cash
payments and Medicaid act as incentives for some parents to coach
and, therefore, they are concerned about the extent to which parents
can manipulate the disability determination process.  However, we
believe that measuring the extent to which coaching may actually
occur is extremely difficult. 

Unless parents admit to it, coaching is almost impossible to
substantiate.  The nature of the parent-child relationship makes
investigating coaching allegations difficult.  Many communications
between parent and child take place at home, out of the view of
outside observers.  In addition, the variability of children's
behavior makes knowing whether a child's behavior is the result of
coaching difficult.  Behavior can vary naturally among children of
the same age--or in the same child over time--as they go through
stages in development or respond to changes in their home or school
environment.  If a child started misbehaving in school, investigators
would need baseline evidence to establish that the child had not
misbehaved extensively in the past.  Finally, even if investigators
could identify a sudden change in behavior, they would have to rule
out other reasons for the change, such as changes in the child's
household or neighborhood environment.  In short, knowing whether the
child is performing poorly or misbehaving because of coaching or for
other reasons is difficult. 

Because coaching is difficult to detect, the extent of coaching
cannot be measured with much confidence.  In recent studies, SSA and
the HHS IG reviewed case files and identified scant evidence of
coaching or malingering.\8 In the rare instances where they found
evidence of possible coaching or malingering, most of the claimants
had been denied benefits anyway.  (App.  III summarizes the results
of the SSA and IG studies, including their scopes and methodologies.)


--------------------
\8 SSA considered possible coaching to be involved in any case in
which the child reported or an information source suspected that the
parent or other caregiver had told the child to misbehave or perform
poorly.  SSA also looked for evidence that the child had malingered;
that is, deliberately provided wrong information or did not put forth
his or her best effort during testing, regardless of whether coaching
was suspected. 


      ACTIONS TO REDUCE PROGRAM'S
      POSSIBLE VULNERABILITY TO
      COACHING
---------------------------------------------------------- Letter :6.1

To protect program integrity, SSA has taken several steps to help
provide assurance that the process can detect coaching or malingering
and then make the appropriate eligibility determination.  In June
1994, SSA began requiring DDSs to report to SSA's regional quality
assurance units any case with an allegation or suspicion of coaching. 
Such cases include those in which teachers, physicians, or
psychologists indicate that (1) the child's behavior was atypical of
the child's customary school behavior, (2) the child was
uncooperative during testing, or (3) the child's behavior
deteriorated without explanation during the 6-month period preceding
the application.  According to SSA, its regional quality assurance
units review all alleged cases of coaching.  As of mid-January 1995,
DDSs nationwide had reported alleged coaching in 674 childhood
cases--or less than one-half of 1 percent of all childhood
applications filed during the period--and fewer than 50 of these
children had been awarded benefits. 

Along with this new requirement, in August 1994, SSA required DDSs to
send applicants' schools a set of questions specifically designed to
elicit the teacher's views on whether the child had been coached. 
Additionally, each SSA regional office has established toll-free
telephone numbers for the exclusive use of teachers and school
officials to notify the regional quality assurance unit of coaching
allegations.  In mid-November 1994, SSA instructed DDSs to begin
distributing these toll-free numbers to schools.  Also, SSA has
instructed its field offices and telephone service centers to report
to the regional quality assurance units any allegations of coaching
received from the general public.  As of mid-January 1995, from all
of these sources, SSA had received a total of 42 telephone calls with
allegations of coaching involving 54 individuals.  According to SSA,
each allegation from teachers, school officials, or the general
public is reviewed if the child was awarded benefits. 


   CONCLUSION
------------------------------------------------------------ Letter :7

Childhood disability decisions based on the IFA process are among the
toughest that DDSs must make.  Particularly in assessing behavioral
and learning disabilities, the level of judgment required makes the
IFA process difficult to administer consistently.  Moreover, the high
level of subjectivity leaves the process susceptible to manipulation
and the consequent appearance that children can fake mental
impairments to qualify for benefits.  Indeed, the rise in allegations
of coaching may reflect public suspicion of a process that has
allowed many children with less severe impairments to qualify for
benefits.  Although scant evidence exists to substantiate that
coaching is a problem, coaching cannot be ruled out and its extent is
virtually unmeasurable. 

We believe that a more fundamental problem than coaching is
determining which children are eligible for benefits using the new
IFA process.  Our analysis documents the many subjective judgments
built into each step of the IFA process to assess where a child's
behavior falls along the continuum of age-appropriate functioning. 
Moreover, studies by SSA and the IG of children awarded benefits for
behavioral and learning disorders illustrate the difficulties that
SSA has experienced over the last 4 years in making definitive and
consistent eligibility decisions for children with these disorders. 

SSA's efforts have been aimed at process improvements rather than
reexamining the conceptual basis for the IFA.  Despite its efforts,
too much adjudicator judgment remains.  Although better evidence and
more use of objective tests where possible would improve the process,
the likelihood of significantly reducing judgment involved in
deciding whether a child qualifies for benefits under the IFA is
remote.  We believe that more consistent decisions could be made if
adjudicators based functional assessments of children on the
functional criteria in SSA's medical listings.  This change would
reduce the growth in awards and target disability benefits toward
children with more severe impairments. 


   MATTER FOR CONSIDERATION BY THE
   CONGRESS
------------------------------------------------------------ Letter :8

Given widespread concern about growth in the SSI program for children
and in light of our findings about the subjective nature of the IFA
process, the Congress could take action to improve eligibility
determinations for children with disabilities.  One option the
Congress could consider is to eliminate the IFA, which would require
amending the statute.  The Congress could then direct SSA to revise
its medical listings, including the functional criteria, so that all
children receive functional assessments

We did not request official agency comments from SSA on a draft of
this report.  However, we discussed the draft with SSA program
officials, who generally agreed that we had accurately characterized
the IFA process and the results of studies.  SSA officials had some
technical comments, which we have incorporated where appropriate. 

Please contact me on (202) 512-7215 if you have any questions about
this report.  Other major contributors are Cynthia Bascetta, Ira
Spears, Ken Daniell, David Fiske, and Ellen Habenicht. 

Jane L.  Ross
Director, Income Security Issues


SCOPE AND METHODOLOGY
=========================================================== Appendix I

To develop the information in this report, we (1) reviewed SSA's
childhood disability program policies, procedures, and records, and
discussed the IFA process with SSA program officials on the national,
regional, and local level; (2) interviewed officials in state DDSs;
(3) reviewed SSA's report on its 1994 study of children with
behavioral and learning disorders; and (4) attended a June 1994 SSA
training course that was based on findings from its study.  We also
discussed eligibility issues with officials of HHS' IG, which
recently issued two reports on the SSI childhood disability
program.\9

To develop SSI childhood program award rate data, we obtained SSA's
computerized records on the results of initial determinations and
reconsideration disability decisions made by DDSs for children under
18 years old from 1988 through September 1994.\10 These records
exclude the results of disability decisions made by administrative
law judges.  From these records, we determined (1) the overall award
rate for children, (2) the percentage of IFA awards that were based
on mental impairments versus physical impairments, (3) the average
monthly number of childhood applications, and (4) the average monthly
number of awards that were based on IFAs versus medical listings. 
These data, as applicable, were determined for the following periods: 
(1) 2 years before the Supreme Court's Sullivan v.  Zebley decision
(Jan.  1, 1988, through Feb.  20, 1990); (2) 2 years after the IFA
process was implemented (Feb.  11, 1991, through Dec.  31, 1992); (3)
January-December 1993; and (4) January-
September 1994.  Because no IFA process existed before the Zebley
decision, no pre-Zebley awards were decided based on IFAs. 

We excluded children who had applied during 1988 through February 10,
1991, from the universe of children on whom decisions were made from
February 11, 1991, through September 30, 1994.  We did this to
minimize the extent to which data in these comparison periods reflect
the result of cases readjudicated as part of the settlement in the
Zebley class action lawsuit.  We were not able to identify or exclude
Zebley classmembers for whom benefits had been denied or terminated
from 1980 through 1987 from any of the comparison periods.  According
to SSA, Zebley classmembers are more likely to have physical
impairments than the general population of new SSI child applicants. 

We performed our work from May 1994 through February 1995 in
accordance with generally accepted government auditing standards. 


--------------------
\9 See Concerns About the Participation of Children with Disabilities
in the Supplemental Security Income Program (A-03-94-02602),
Department of Health and Human Services, Office of Inspector General
(Oct.  13, 1994); and Supplemental Security Income:  Disability
Determinations for Children with Mental Impairments (A-03-94-02603),
Department of Health and Human Services, Office of Inspector General
(Jan.  26, 1995). 

\10 The childhood program statistics presented in this report were
developed using the same basic methodology used in Social Security: 
Rapid Rise in Children on SSI Disability Rolls Follows New
Regulations (GAO/HEHS-94-225, Sept.  9, 1994).  This report focused
on the growth in awards after SSA changed the disability criteria for
children. 


EFFORTS TO AFFECT IMPLEMENTATION
OF THE ZEBLEY DECISION
========================================================== Appendix II


   JANUARY 1991
-------------------------------------------------------- Appendix II:1

One month before SSA issued regulations implementing the new IFA
process, the Zebley plaintiff's counsel submitted interrogatories to
SSA asking, among other things, why nine DDSs with the lowest award
rates for children had such low award rates.\11

SSA regional officials were tasked with answering some of the
counsel's interrogatories and, in some instances, the officials
informed the states that they were the subject of the counsel's
inquiry.  Also, from time to time thereafter, SSA officials shared
state-by-state award rate data with state DDSs.  Some SSA regional
officials stated that they believed some DDSs could have felt
pressured to increase their award rates. 


--------------------
\11 The nine states were Alabama, Arkansas, Colorado, Louisiana,
Mississippi, Nebraska, South Carolina, West Virginia, and Wisconsin. 


   FEBRUARY 1991
-------------------------------------------------------- Appendix II:2

In the month that SSA issued regulations implementing the new IFA
process, a federal district court ordered SSA to perform special
quality assurance reviews of disability applications denied under the
new regulations.  The court order required SSA to do quality
assurance reviews of denials made by 10 state DDSs that, according to
SSA, Zebley plaintiff's counsel had identified as denial prone due to
their low award rates.\12 Based on its own studies, SSA had argued
before the court that low award rates were not reliable indicators of
whether special corrective action was needed to avoid incorrect
denials, but the court required SSA to implement the special quality
assurance reviews for these 10 states. 

Under the court order, during the first month after the new
regulations were in effect, SSA had to review the lesser of 100 or
all denials for each denial-prone state.  SSA reviewed only 25
denials for other states.  A subsequent March 1991 court order
required SSA, after the first month, to review at least 1,000 denials
per month nationwide.  SSA's sample of 1,000 denials included 15
percent of the denials from each of the 10 denial-prone states. 

By memorandum in February 1991, SSA informed all DDSs of the special
quality assurance requirements and identified the 10 states that had
been classified as denial prone.  The court order required that SSA
send the results of the quality assurance reviews monthly to the
Zebley plaintiff's counsel. 


--------------------
\12 The 10 states were Alabama, Arkansas, Colorado, Louisiana,
Mississippi, Nebraska, New Mexico, South Carolina, West Virginia, and
Wisconsin. 


   JUNE 1991
-------------------------------------------------------- Appendix II:3

The Zebley plaintiff's counsel wrote to the SSA Commissioner citing a
"disturbing pattern" of low allowance rates in eight states and asked
the Commissioner to take remedial steps.\13


--------------------
\13 The eight states were Connecticut, Kentucky, Louisiana, Nebraska,
New Mexico, Texas, West Virginia, and Wisconsin. 


   DECEMBER 1992
-------------------------------------------------------- Appendix II:4

In a newsletter to legal aid societies, the Zebley counsel listed 13
DDSs whose cumulative allowance rates were at 50 percent or below.\14
The counsel encouraged legal aid society representatives in those
states to contact the DDS directors and "confront them with their
sub-par performance."

APPENDIX III

--------------------
\14 The 13 states were Arkansas, Connecticut, Louisiana, Maine,
Mississippi, Missouri, Montana, Nebraska, New Mexico, South Carolina,
Tennessee, Texas, and West Virginia. 


STUDIES DONE BY SSA AND THE
INSPECTOR GENERAL
========================================================== Appendix II


   1994 STUDY BY SSA'S OFFICE OF
   DISABILITY
-------------------------------------------------------- Appendix II:5

SSA considers behavioral and learning disorders to be the most
susceptible to coaching and malingering.  In 1994, SSA's Office of
Disability in Baltimore reviewed a national sample of 617 school-age
children who had applied due to behavioral and learning disorders. 
Because the sample was small, the findings of the study cannot be
projected to the universe of childhood disability claims or to the
subset of specific impairments studied. 


      SCOPE AND METHODOLOGY
------------------------------------------------------ Appendix II:5.1

The 617 children were selected from those who had applied due to such
impairments as attention deficit disorder, attention deficit
hyperactivity disorder, personality disorder, conduct disorder,
learning disorder, oppositional defiant disorder, anxiety disorder,
developmental delay, behavior disorder, speech and language
disorders, borderline intellectual functioning, and adjustment
disorder.  According to SSA, these types of disorders constitute
about 20 percent of all childhood disability applications.  SSA
excluded cases involving extremely severe mental disorders, such as
psychotic disorders and mental retardation. 

SSA selected the 617 cases from final DDS decisions that SSA's
regional quality assurance staff had already reviewed for accuracy. 
The 617 cases in the sample consisted of 325 awards and 292 denials
that DDSs adjudicated during October 1992 through July 1993.  SSA
reviewed case file documentation for the 617 cases. 


      COACHING
------------------------------------------------------ Appendix II:5.2

In its review of case file documentation, SSA considered coaching to
be involved in any claim in which the child reported or an
information source suspected that the parent or other caregiver had
told the child to act or respond in a manner that would make the
child appear more functionally limited than he or she actually was. 
In addition, SSA looked for evidence indicating that the child had
malingered; that is, deliberately provided wrong information or did
not put forth his or her best effort during testing. 

SSA found only 13 cases that showed any evidence of possible coaching
or malingering, and only 3 of these cases were awards.  In all cases,
the evidence indicating possible coaching was provided by medical
professionals or psychologists who performed consultative
examinations for SSA.  None of the evidence indicating possible
coaching or malingering was provided by schools.  The three
questioned awards involved children who may have malingered during IQ
testing.  In these cases, however, the awards were based on factors
other than the results of the testing.  For example, one child with
an oppositional defiant disorder appeared to malinger during IQ
testing administered by a consultative examiner, but the award was
based on other problems stemming from the disorder, not the results
of the testing. 


      INCORRECT AWARDS
------------------------------------------------------ Appendix II:5.3

Of the 325 awards reviewed by SSA, SSA found that 8.6 percent (28)
should have been denials and another 27.7 percent (90) should not
have been made without obtaining more supporting documentation.  We
asked SSA, based on experience in its quality assurance program, to
estimate how many of the 90 cases with insufficient documentation
would have been denials if all documentation had been obtained, and
SSA estimated that 13 (or 4 percent of the 325 awards) would have
been denials.  Thus, we concluded that a total of 41 awards (12.6
percent of the 325 awards) should have been denials.  By contrast, of
292 denials reviewed in the study, SSA found that only 1.4 percent
(4) should have been awards, and another 1.4 percent (4) should not
have been made without obtaining more supporting documentation. 

Combining all decisional and documentational errors for the 617
denials and awards in SSA's study, the overall error rate for this
group of cases was 20.4 percent.\15 This is about twice the maximum
acceptable error rate of 9.4 percent that SSA allows for decisional
and documentational errors combined for all initial disability
decisions made by an individual DDS. 

According to SSA's Office of Disability, a primary reason that DDSs
made awards that should have been denials was that DDSs had
frequently overrated--but rarely underrated--the severity of
children's functional limitations.  Such overrating occurred
primarily because DDSs had (1) compared the child with the perfect
child rather than the average child, (2) based the limitation on a
single incident rather than behavior over time, (3) not considered
the child's ability to function while on an effective medication
regimen, and (4) based the limitation on the child's life
circumstances rather than the effects of a medically determinable
impairment. 

DDSs also had mechanically applied SSA's guidelines on how to make
awards using the results of the IFA process.  SSA's guidelines
instruct DDSs that they generally should award benefits to children
who have moderate limitations in any three of the areas of ability
assessed in the IFA process.  SSA found, however, that DDSs had used
this instruction as a rule rather than a guideline.  DDSs had
automatically made awards to any child with three moderate
limitations, regardless of how strong or weak the moderate
limitations were.  SSA stated that its guideline assumed "three good,
solid moderates." SSA found that, when DDSs had identified two
moderate limitations, they sometimes made special attempts to find a
third moderate limitation even though the evidence did not support
it. 

DDSs had also "double-weighed" the effects of impairments in more
than one of the areas of ability assessed in the IFA process, making
the impairment seem more severe and pervasive than it actually was. 
For example, in some cases children displayed a lack of self-control
by exhibiting more than one inappropriate behavior, such as fighting,
aggressive behavior, disrespectful behavior, lying, oppositional
behavior, and stealing.  Although all these behaviors should have
been rated only in the personal/behavioral area, DDSs had rated some
behaviors in the personal/behavioral area and others in the social
abilities area, giving the child moderate limitations in two areas
rather than only one.  This meant that the child needed only one more
moderate limitation to have the three moderate limitations needed for
approval. 

SSA also found that DDSs had sometimes based decisions on old
evidence when current evidence indicated children had improved and
that DDSs had sometimes assessed limitations that could not be
attributed to medical impairments. 


--------------------
\15 The overall error rate for the 617 cases was computed as follows: 
(28 award decisional errors + 90 award documentational errors + 4
denial decisional errors + 4 denial documentational errors)/ 617 =
0.204. 


   INSPECTOR GENERAL STUDY
-------------------------------------------------------- Appendix II:6

As the IG reported in January 1995, IG staff reviewed the case files
for a sample of 553 children whose applications were adjudicated by
DDSs in 1992.  Of the 553 children, 298 had been awarded benefits by
10 DDSs--Connecticut, Illinois, Kentucky, New York, North Carolina,
North Dakota, Pennsylvania, South Dakota, Vermont, and Wisconsin. 
The remainder of the 553 cases consisted of a nationwide sample of
255 denials.  Of the 298 awards, 129 (43 percent) had been decided
based on an IFA, and 195 of the 255 denials (76 percent) had been
decided based on an IFA.  The IG targeted its study at cases
involving mental retardation, attention deficit hyperactivity
disorder, and other learning and behavioral disorders.  Based on its
review of these cases, IG officials told us that they had found no
evidence of coaching. 

As the IG reported, when the IG staff had questions about the
accuracy of a DDS disability determination or about the sufficiency
of the evidence supporting a determination, the IG provided the case
file to SSA's Office of Disability in Baltimore--the same staff
responsible for conducting SSA's study of 617 childhood disability
claims.  The Office of Disability reviewed the accuracy of each of
the questioned cases.  The IG staff also visited the 10 DDSs to
obtain their opinions on the adequacy of the SSA guidelines used to
make disability determinations. 

Of the 129 awards reviewed that were based on IFAs, the IG reported
that 17 (13 percent) should have been denials and another 38 (29
percent) were based on insufficient evidence.  The IG attributed this
problem to DDSs having difficulty in interpreting and complying with
SSA guidelines for obtaining and evaluating evidence concerning the
severity of the mental impairments of children on whom IFAs are
conducted.  The IG stated that these children have less severe
impairments than those children determined to be disabled based on
the impairment listing, making the assessment of the effects of their
impairments on their ability to function age-appropriately more
difficult.  In discussions with employees of the 10 DDSs, the IG
reported that many expressed concern that the SSA guidelines for
determining disability for children with mental impairments were not
sufficiently clear or objective. 

