Disability Insurance: SSA Should Strengthen Its Efforts to Detect
and Prevent Overpayments (10-SEP-04, GAO-04-929).		 
                                                                 
The Social Security Administration's (SSA) Disability Insurance  
(DI) program is one of the nation's largest cash assistance	 
programs for disabled workers. In fiscal year 2003, the DI	 
program provided about $70 billion in financial assistance to	 
approximately 7.5 million disabled workers, their spouses, and	 
dependent children. This program has grown in recent years and is
poised to grow further as the baby boom generation ages. The	 
Senate Committee on Finance asked GAO to (1) determine the amount
of overpayments in the DI program, particularly those		 
attributable to earnings or work activity, and (2) identify any  
vulnerabilities in SSA's processes and policies for verifying	 
earnings that may contribute to work-related overpayments.	 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-04-929 					        
    ACCNO:   A12189						        
  TITLE:     Disability Insurance: SSA Should Strengthen Its Efforts  
to Detect and Prevent Overpayments				 
     DATE:   09/10/2004 
  SUBJECT:   Beneficiaries					 
	     Data integrity					 
	     Debt						 
	     Disability benefits				 
	     Disability insurance				 
	     Federal employee disability programs		 
	     Management information systems			 
	     Overpayments					 
	     Policy evaluation					 
	     Policies and procedures				 
	     SSA Disability Insurance Program			 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO 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.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-04-929

United States Government Accountability Office

                    GAO Report to the Chairman, Committee on
                              Finance, U.S. Senate

September 2004

                                   DISABILITY
                                   INSURANCE

      SSA Should Strengthen Its Efforts to Detect and Prevent Overpayments

GAO-04-929

Highlights of GAO-04-929, a report to the Chairman, Committee on Finance,
U.S. Senate

The Social Security Administration's (SSA) Disability Insurance (DI)
program is one of the nation's largest cash assistance programs for
disabled workers. In fiscal year 2003, the DI program provided about $70
billion in financial assistance to approximately 7.5 million disabled
workers, their spouses, and dependent children. This program has grown in
recent years and is poised to grow further as the baby boom generation
ages. The Senate Committee on Finance asked GAO to (1) determine the
amount of overpayments in the DI program, particularly those attributable
to earnings or work activity, and (2) identify any vulnerabilities in
SSA's processes and policies for verifying earnings that may contribute to
work-related overpayments.

GAO is making recommendations to the Commissioner of Social Security
directing the agency to explore new tools and data sources that can be
used to more effectively detect and prevent earnings-related overpayments.
SSA agreed with GAO's recommendations and provided information on several
initiatives that are planned or underway to address them, such as a new
computer match using information from the Office of Child Support
Enforcement's National Directory of New Hires to verify beneficiaries'
earnings in a more timely manner.

September 2004

DISABILITY INSURANCE

SSA Should Strengthen Its Efforts to Detect and Prevent Overpayments

Overpayment detections in the DI program increased from $772 million in
fiscal year 1999 to about $990 million in 2003. The true extent of
overpayments resulting from earnings that exceed agency guidelines is
currently unknown. Based on available data from SSA, GAO found that about
31 percent of all DI overpayments are attributable to DI beneficiaries who
worked and earned more than allowed. Moreover, GAO found that these
overpayments contributed to mounting financial losses in the program. From
1999 to 2003, total overpayment debt increased from about $1.9 billion to
nearly $3 billion.

Three basic weaknesses impede SSA's ability to prevent and detect
earningsrelated overpayments. First, the agency lacks timely data on
beneficiaries' earnings and work activity. Second, SSA uses inefficient
processes to perform work continuing disability reviews (work CDRs).
Third, the agency relies on potentially inaccurate management information
to effectively monitor and oversee some parts of this workload. These
weaknesses contributed to some work CDR cases GAO identified that were as
much as 7 years old, resulting in potential and established overpayments
as large as $105,000 per beneficiary. In addition, GAO found that SSA
relies on potentially inaccurate management information to administer its
work CDR workload. SSA is developing new automated systems that may
potentially address some of these problems and could help the agency
balance the important goals of encouraging individuals with disabilities
return to work, while also ensuring program integrity. However, it is too
early to determine how effective such systems will be.

Total Overpayment Debt Is Increasing (1999-2003)

Dollars in millions 3,000

2,500

2,000

1,500

1,000

500

0 1999 2000 2001 2002 2003 Year

Source: GAO analysis based on SSA data.

www.gao.gov/cgi-bin/getrpt?GAO-04-929.

To view the full product, including the scope and methodology, click on
the link above. For more information, contact Robert E. Robertson at (202)
512-7215 or [email protected].

Contents

  Letter

Results in Brief
Background
Overpayments in the DI Program Are Substantial and Have

Increased in Recent Years

Lack of Timely Data on Beneficiary Earnings and Other Vulnerabilities
Impede SSA's Ability to Detect and Prevent DI Overpayments

Conclusions
Recommendations
Agency Comments and Our Evaluation

                                       1

                                      3 4

                                       7

10 17 18 19

Appendix I Scope and Methodology

Appendix II Comments from the Social Security Administration

Appendix III GAO Contacts and Staff Acknowledgment 31

GAO Contacts 31 Staff Acknowledgments 31

Related GAO Products

  Figures

Figure 1: Total DI Overpayment Detections Have Increased (1999- 2003) 8
Figure 2: Total Overpayment Debt Is Increasing (1999-2003) 9

Abbreviations

CDR continuing disability review
DDS Disability Determination Service
DI Disability Insurance
IRS Internal Revenue Service
NDNH National Directory of New Hires
OCO Office of Central Operations
OCSE Office of Child Support Enforcement
OIG Office of Inspector General
PSC Program Service Center
SSA Social Security Administration
SGA substantial gainful activity
SSI Supplemental Security Income

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States Government Accountability Office Washington, DC 20548

September 10, 2004

The Honorable Charles E. Grassley Chairman, Committee on Finance United
States Senate

Dear Mr. Chairman:

The Social Security Administration's (SSA) Disability Insurance (DI)
program is one of the nation's largest cash assistance programs for
disabled workers. In fiscal year 2003, the DI program provided about $70
billion in financial assistance to approximately 7.5 million disabled
workers, their spouses, and dependent children. This program has grown in
recent years and is poised to grow further as the baby boom generation
ages. Given the concerns about the long-term solvency of the DI trust
fund, it is important for SSA to ensure that only truly eligible
individuals receive benefits.

SSA guidelines state that DI beneficiaries are permitted to earn up to
$810 per month in calendar year 2004-a level of earnings referred to as
substantial gainful activity (SGA)1-for a limited period of time without
losing eligibility for benefits. After completing a 9-month "trial work
period," beneficiaries who earn more than SGA are generally ineligible for
future DI payments, and may be overpaid if SSA does not stop their
benefits in a timely manner. The potential of having to repay a large
overpayment may discourage some beneficiaries from continuing to work,
thus running contrary to SSA's goal of helping such individuals become
self-sufficient. SSA conducts continuing disability reviews (CDR) of
beneficiaries' earnings and work activity to determine whether a claimant
remains financially eligible for DI benefits.2 The agency refers to these
reviews as "work CDRs." These reviews generally require SSA staff to
perform several steps to assess beneficiaries' continuing eligibility for
benefits, including mailing notices to beneficiaries requesting
information about their work activity, contacting the beneficiaries'
employer(s) to

1The SGA level changes annually. For example, SGA in calendar year 2003
was $800 per month.

2SSA also conducts "medical" CDRs to evaluate whether a beneficiary has
medically improved to the point where they are able to work.

verify their monthly earnings, and assessing several variables that can
affect eligibility, including employer subsidies and work-related
expenses.

Given the importance of verifying DI beneficiaries' earnings to ensure
they receive the correct amount of benefits, the Senate Committee on
Finance asked us to (1) determine the amount of overpayments in the DI
program, particularly those attributable to earnings or work activity, and
(2) identify any vulnerabilities in SSA's processes and policies for
verifying earnings that may contribute to work-related overpayments. To
answer these questions, we used an approach similar to the methodology in
our prior reviews of SSA's Supplemental Security Income (SSI) program.3 In
particular, we reviewed DI performance data, prior reports by SSA and its
Office of Inspector General (OIG), external research studies, and our
prior reviews on the program. We analyzed DI payment data over a 5-year
period from 1999 to 2003 and examined between 5 and 7 work CDR cases from
each of the SSA field offices and program service centers we visited. In
addition, we randomly selected and reviewed 71 work CDR cases from one of
SSA's program service centers to determine if they were processed in
accordance with program guidelines. Finally, we conducted in-depth
interviews with 230 management and line staff from SSA's headquarters; its
regional offices in New York and San Francisco; 18 field offices in 6
states-California, Florida, Maryland, Massachusetts, New York, and
Virginia; and 3 out of 8 regional program service centers.4 During our
meetings, we (1) examined existing work CDR procedures; (2) documented
management and staff views on the effectiveness of SSA's work CDR
processes for detecting and preventing earnings-related overpayments; and
(3) discussed potential improvements to existing program processes,
systems, and policies. We assessed the reliability of all databases used
in our review, and found them to be sufficiently reliable for the purposes
of this report. See appendix I for details on the scope and methodology of
our review. We performed our work from September 2003 through June 2004 in
accordance with generally accepted government auditing standards.

3We designated SSI a high-risk program in 1997 after several years of
reporting on specific instances of abuse and mismanagement, increasing
overpayments, and poor recovery of outstanding debt. SSA subsequently made
several changes to improve SSI program integrity. We removed SSI from our
high-risk list in 2003.

4The program service centers (PSC) are responsible for a variety of
activities, including work CDRs.

  Results in Brief

Overpayment detections in the DI program increased from about $772 million
in fiscal year 1999 to about $990 million in 2003. The true extent of
overpayments resulting from earnings that exceed agency guidelines is
currently unknown, but could be higher than available data indicate. On
the basis of data from a recent SSA study, we estimate that about 31
percent of all DI overpayments are attributable to DI beneficiaries who
worked and earned more than SGA. Moreover, we found that these
overpayments contributed to mounting overpayment debt, which increased
from about $1.9 billion to nearly $3 billion during the same period.
Although SSA increased overpayment collections during this time, our
analysis shows that overpayment waivers (overpayments that SSA decides not
to collect) and write-offs (overpayments that SSA determines cannot be
collected) also increased. Thus, financial losses are mounting,
contributing to a widening gap between total overpayment debt and annual
overpayment collections.

Three basic weaknesses impede SSA's ability to prevent and detect
earnings-related overpayments: The agency (1) lacks timely data on
beneficiaries' earnings and work activity, (2) uses inefficient processes
to perform work CDRs, and (3) relies on potentially inaccurate management
information to manage a portion of its work CDR workload. First, SSA's
main source of earnings verification for the DI program is derived from
matching its own earnings database with Internal Revenue Service (IRS)
wage data, which is typically 12-18 months old when it is first available
to SSA. The agency does not currently have the authority to conduct
computer matches with the Office of Child Support Enforcement's National
Directory of New Hires-a database with more timely wage information.
Second, SSA uses inefficient processes to perform work CDRs. In
particular, the agency lacks an effective screen to help identify cases
most likely to result in large overpayments. Nor does the agency have an
automated alert system that could notify field offices and program service
centers about cases at high-risk for overpayments. Finally, SSA relies on
potentially inaccurate management information to effectively administer
its CDR workload. In particular, its data may not accurately reflect the
age and disposition of its work CDR workload, or the time it actually
takes to process them. These vulnerabilities may contribute to the "old"
cases we identified in many SSA field offices, some of which were as much
as 7 years old, resulting in large individual overpayments totaling
between $28,000 and $105,000. Moreover, we found that SSA has difficulty
balancing competing workloads-particularly in its field offices where
staff resources are limited and staff have numerous different duties-that
may contribute to some of the old cases we observed. SSA is developing new
automated systems that could potentially address some of these

problems by helping the agency manage its disability workload more
efficiently, but it is too early to determine if these initiatives will
address the weaknesses we identified.

Work CDRs can be complex and time-consuming for SSA staff to perform. We
recognize that ensuring program integrity while focusing on the important
goal of returning individuals with disabilities to work presents
additional challenges for SSA. However, there are several areas where we
believe SSA can make improvements. Accordingly, we are recommending that
the Commissioner of Social Security direct the agency to explore new tools
and more timely data sources that can be used to more effectively detect
and prevent earnings-related overpayments.

SSA agreed with our recommendations and provided information on several
initiatives that are planned or underway to address them.

Background 	The DI program was established in 1956 to provide monthly cash
benefits to individuals who were unable to work because of severe
long-term disability. In fiscal year 2003, SSA paid about $70 billion to
7.5 million disabled workers, their spouses, and dependents, with average
monthly cash benefits of about $723 per beneficiary.5 To be eligible for
benefits, individuals with disabilities must have a specified number of
recent work credits under Social Security when they first became disabled.
Individuals may also be able to qualify based on the work record of a
deceased, retired, or disabled parent, or a deceased spouse. Benefits are
financed by payroll taxes paid into the Federal Disability Insurance Trust
Fund by covered workers and their employers, based on the worker's
earnings history. To meet the definition of disability under the DI
program, an individual must have a medically determinable physical or
mental impairment that (1) has lasted or is expected to last at least 1
year or to result in death and (2) prevents the individual from engaging
in substantial gainful activity. Individuals are engaged in SGA if they
have earnings above $810 per month in calendar year 2004.6 Program
guidelines require DI beneficiaries to report their earnings to SSA in a
timely manner in order to ensure that they remain eligible for benefits.

5Average benefit figure was reported for December 2003. 6SGA for blind
beneficiaries is $1,350 per month.

SSA conducts work issue CDRs to determine if beneficiaries are working
above the SGA level.7 SSA initiates a work CDR only after the beneficiary
has completed a 9-month trial work period, during which the beneficiary is
allowed to earn more than the SGA level without affecting their
eligibility for benefits.8 The trial work period is one of several
provisions in the DI program intended to encourage beneficiaries to return
to work. The trial work period begins with the first month a beneficiary
is eligible for DI benefits. Once the trial work period is completed,
beneficiaries are generally ineligible for future DI benefits unless their
earnings fall below the SGA level.9

Work CDRs are triggered by several types of events, although most are
generated by SSA's Continuing Disability Review Enforcement Operation
(enforcement operation). This process involves periodic computer matches
between SSA's administrative data and IRS wage data. The enforcement
operation generates notices for cases that exceed specified earnings
thresholds,10 which are forwarded to 1 of 8 program service centers for
additional examination.11 The cases at each program service center are
then temporarily housed in a central repository (called the computer
output section) and are released to "earnings reviewers" for

7We use the term "work CDRs" to describe instances in which SSA staff
perform limited development of beneficiary earnings because they determine
that a full work CDR is not necessary (an activity that SSA refers to as a
"work CDR action"), as well as "full" work CDRs in which a case is fully
developed and staff fill out specific forms to receive work credit for
completing a work CDR.

8The trial work period allows beneficiaries to work for 9 months (not
necessarily consecutive) within a 60-month rolling period during which
they may earn any amount without affecting benefits.

9To provide additional incentives to encourage work, beneficiaries who
have completed their trial work period are entitled to a 36-month extended
period of eligibility during which they may receive benefits for any month
in which their earnings fall below SGA. Other work incentive provisions
allow SSA to deduct certain impairment-related work expenses and employer
subsidies from beneficiaries' earnings determination. SSA staff must
consider all these provisions when assessing whether beneficiaries'
earnings constitute SGA.

10SSA currently uses six times the SGA amount, or $4,860 as the annual
test level to screen out beneficiaries whose earnings amount would not
likely affect their DI benefits.

11Most cases (about 60 percent) are sent to the program center in SSA's
Office of Central Operations (OCO). OCO is responsible for handling
beneficiaries who are less than 55 years of age. According to SSA
officials, these beneficiaries tend to work more frequently and have more
employers than older beneficiaries, thus making the cases more complicated
to process. The remaining cases for beneficiaries older than 55 are sent
to one of the remaining 7 program service centers.

development of work activities. Cases are generally released for
development on a first-in-first-out basis, based on how long they have
been in the central repository, and according to staff workloads. After
initial review, cases for which individuals may require cessation of
benefits are generally forwarded to a "disability processing specialist"
for additional development.12

Work CDRs can also be triggered by other events. For example, SSA requires
beneficiaries to undergo periodic medical examinations to assess whether
they continue to be physically disabled.13 During such reviews, Disability
Determination Service staff sometimes discover evidence that indicates the
beneficiary may be working and usually forwards the case to an SSA field
office or program service center for earnings/work development. Additional
events that may trigger a work CDR include reports from state vocational
rehabilitation agencies, other federal agencies, and anonymous tips.
Finally, DI beneficiaries may voluntarily report their earnings to SSA by
visiting an SSA field office, or calling the agency's toll free "800"
number.

Several SSA components are involved in processing work CDRs. While most
are initially sent to SSA's program service centers as a result of the
enforcement operation, some cases are referred to any one of SSA's more
than 1,300 field offices for more in-depth development. Field offices also
tend to be the focal points for work CDRs generated by events other than
the enforcement operation. Work CDRs can entail labor-intensive,
timeconsuming procedures such as reviewing folders, performing in-person
interviews, and contacting beneficiaries and their employers to verify
their monthly earnings. Staff are also required to take into consideration
several complex work incentive provisions when calculating whether
earnings

12"Disability processing specialists" work in SSA's program service
centers and are responsible for determining if benefits should be
discontinued and whether an overpayment exists. "Earnings reviewers" in
the program centers are generally responsible for initial analysis of a
beneficiary's earnings; however, only disability processing specialists
have the authority to cease benefits. In SSA's field offices, the claims
representatives are responsible for the duties performed by both the
disability processing specialist and the earnings reviewer.

13SSA contracts with state Disability Determination Services (DDS) that
are responsible for assessing whether an individual is medically disabled
(a "medical" CDR). During the course of a medical CDR, DDS examiners
sometimes find evidence that a beneficiary may be working. Medical CDRs
are costly to perform and such cases are typically referred to an SSA
field office or PSC for financial development before additional medical
development is performed.

exceed SGA.14 In addition, staff-particularly in SSA field offices-are
also required to balance numerous competing workloads, including
processing initial claims, serving individuals who walk into the field
office without an appointment, meeting with beneficiaries who have
requested an appointment, and processing a "special disability
workload."15

  Overpayments in the DI Program Are Substantial and Have Increased in Recent
  Years

DI overpayment detections increased from about $772 million to about $990
million between fiscal years 1999 and 2003. These overpayments included a
substantial amount due to beneficiaries who worked and earned more than
SGA. Our analysis of available overpayment data shows that, on average,
beneficiaries with earnings over program guidelines constitute about 31
percent of all DI overpayments. These overpayments also contributed to
mounting financial losses in the DI program. Total overpayment debt
increased from about $1.9 billion to nearly $3 billion from fiscal years
1999 to 2003. SSA overpayment collections increased from about $269
million to about $431 million during the same period. However, our
analysis shows that waivers and write-offs also increased during this
period.

    Overpayments Due to Beneficiary Earnings above SGA May Be More Prevalent
    than SSA Currently Detects

Total DI overpayment detections increased from about $772 million to about
$990 million between fiscal year 1999 to 2003 (see fig. 1) including a
substantial proportion due to beneficiary earnings.16 On the basis of data
in a recent study from SSA, we calculated that overpayments attributable
to work and earnings averaged about 31 percent of all DI overpayments
annually between 1999 and 2002. We consulted SSA officials about our
calculations to determine if they were accurate. These officials agreed
that the estimate is generally accurate based on limited available data,
but likely understates the true extent of the problem. In particular, SSA
officials acknowledged that their study only examined beneficiaries who
had their benefits suspended or terminated following a work CDR; it did

14These provisions include tracking the 9-month "trial work period" and an
"extended period of eligibility," as well as calculating "impairment
related work expenses" and "employer subsidies."

15This workload is comprised of about 500,000 SSI recipients who at some
point became eligible for DI. However, the SSI administrative systems
failed to identify these cases. SSA is now focusing resources on
processing this workload.

16Overpayments may also be caused by other types of events, including
receipt of workers compensation benefits, being in prison while receiving
benefits, and medical improvement to the point where the individual is no
longer disabled.

not consider individuals who may have been overpaid but continued to
receive benefits. A beneficiary may be overpaid, but not placed in
suspended or terminated status because (1) SSA waived the overpayment, (2)
the case was still being processed, or (3) the individual became
unemployed and returned to the DI rolls. Our review identified several
such cases in numerous field offices. For example, one case we examined
involved a beneficiary who was selected for review by the enforcement
operation every year from 1998 to 2001. Other than notations on the
individual's account that the case was selected for review, there was no
evidence that a work CDR was ever conducted. In February 2003, program
service center staff transferred the case to a field office to have the
recipient's earnings reviewed. However, field office staff were unable to
contact the recipient and the case was transferred back to the program
service center in August 2003. As of March 2004, the case was still being
reviewed and waiting final SSA action. SSA officials told us that this
individual should have had an overpayment listed for the time between
December 1999 and September 2001. However, at the time of our review, no
overpayment had yet been established and, therefore did not appear in
SSA's overpayment detection data for those years. Ultimately, we estimate
that this case will likely result in a $64,000 overpayment once it is
fully developed and completed.

Figure 1: Total DI Overpayment Detections Have Increased (1999-2003)
Dollars in millions

                                      990

1,000

900

800

700

600

500

400

300

200

100

0 1999 2000 2001 2002 2003 Year

Source: GAO analysis based on data from SSA's Office of Finance,
Assessment, and Management.

    Although SSA Has Improved Its Collection Efforts, Financial Losses Are
    Mounting

The increase in DI overpayments from 1999 to 2003 has contributed to
mounting financial losses in the program. Total DI overpayment debt17
increased from about $1.9 billion in 1999 to nearly $3 billion in 2003.
During this same period, SSA's overpayment collections increased from
about $269 million to about $431 million. Agency officials attributed the
increase in collections in part to new initiatives they have made use of.
For example, SSA has conducted debt management workshops to (1) develop
new ideas on collecting the agency's mounting outstanding debt and (2)
identify and prioritize debt that the agency should concentrate on
collecting. In addition, SSA is in the process of developing new
collection tools, such as wage garnishment to recoup overpayments, and has
published final regulations to implement this tool.18 However, these
improvements notwithstanding, the total overpayment debt is increasing.
(See fig. 2.)

Figure 2: Total Overpayment Debt Is Increasing (1999-2003)

Dollars in millions 3,000

2,500

2,000

1,500

1,000

500

0 1999 2000 2001 2002

Year

Source: GAO analysis based on SSA data.

17Total overpayment debt is comprised of existing debt carried forward
from prior years, and newly detected overpayments, net of collections,
waivers, and write-offs in each fiscal year.

18 68 Fed. Reg. 74117 (to be codified at 20 C.F.R. pt. 422 subpt. E ).

Increases in waivers and write-offs19 during this period have also
contributed, in part, to the DI program's growing overpayment debt. SSA
must waive collection of an overpayment if SSA determines that the
beneficiary was not at fault in causing the overpayment and either the
beneficiary would be financially unable to repay the overpayment or
recovery would be against equity and good conscience. The agency may also
write-off overpayments for various reasons, including when the agency is
unable to locate an individual for a prolonged period of time. Waivers and
write-offs increased from about $222 million in 1999 to about $325 million
in 2003. The increase in waivers and write-offs is attributable, in part,
to increases in total program outlays during this period.20 Ultimately,
our review suggests that overpayments not only contribute to increasing
overpayment debt, but also may be a disincentive for individuals with
disabilities to return to work. In particular, the potential of having to
repay a large overpayment may discourage some beneficiaries from
continuing to work, thus running contrary to SSA's goal of helping such
individuals become self-sufficient.

  Lack of Timely Data on Beneficiary Earnings and Other Vulnerabilities Impede
  SSA's Ability to Detect and Prevent DI Overpayments

SSA's ability to detect and prevent earnings-related overpayments is
hindered by a lack of timely wage data, inefficient processes for
conducting work CDRs, and potentially inaccurate management information.
First, the earnings data produced by the enforcement operation are
typically 12-18 months old when SSA first receives it, thus making some
overpayments inevitable. Second, SSA lacks the means to systematically
screen and identify beneficiaries most likely to incur large overpayments.
Moreover, even if such a screen existed, SSA currently lacks an automated
alert mechanism for notifying its field office and program service center
staff about such cases. Third, SSA relies on management information data
that may not accurately reflect the age of

19 According to SSA, some waivers are beyond the agency's control, such as
those attributable to bankruptcy, Tax Refund Offsets, and Administrative
Law Judge decisions instructing the agency to waive overpayments.
Moreover, some debt that is written-off may ultimately be reestablished if
the beneficiary returns to the DI rolls. In addition, SSA continues
recovery efforts of qualified written-off debts via the Treasury Offset
Program, credit bureau reporting, and mandatory cross-program recovery.

20 Total DI program outlays increased from about $50.4 billion in fiscal
year 1999 to about $70 billion in fiscal year 2003.

work CDR cases-the time it actually takes to review and complete them.
Inaccurate management data can impede the agency's ability to effectively
monitor program activities and make corrections, when necessary. These
weaknesses may contribute to some cases becoming old and resulting in
large overpayments. We identified several cases in which as much as 7
years had passed between the point at which the case was initially
selected for development and the time it was completed.

    SSA Lacks Timely Data to Detect Overpayments

SSA currently relies on outdated information to verify DI beneficiaries'
eligibility for benefits.21 The agency conducts periodic matches between
its earnings records and IRS wage data to determine if beneficiaries have
earnings above the SGA level. The Continuing Disability Review Enforcement
Operation (enforcement operation) is generally conducted three times
annually-a principal match in May, and two supplemental matches in August,
and February of the following year. According to some SSA officials,
earnings data from the enforcement operation are generally about 12-18
months old by the time the cases are selected for review and arrive in the
program service center. SSA officials told us that the age of the earnings
data impedes the agency's ability to effectively detect potential
overpayments in a timely manner. Moreover, because a substantial
proportion of all work CDRs in any given year are generated by these
enforcement matches, a large proportion of this workload is dependent on
outdated earnings information. Thus, some cases with potentially large
overpayments may not be detected for extended periods of time.

SSA lacks access to more timely sources of wage data for verifying DI
beneficiaries' earnings, such as the Office of Child Support Enforcement's
National Directory of New Hires (NDNH). This database contains quarterly
state wage and new hires data that could be used to help evaluate
beneficiaries' continuing eligibility for benefits more quickly than the
enforcement operation. While SSA currently uses this database to
periodically monitor the earnings of SSI recipients, it lacks similar
authority for the DI program. In particular, SSA currently lacks the

21Beneficiaries are required to report earnings to SSA that may affect
their eligibility for benefits, and SSA relies on beneficiaries to report
such information in a timely manner. However, our review found that
individuals sometimes do not report their earnings as required.

authority to conduct "batch file" computer matches22 with the NDNH-similar
to the types of matches it routinely uses to verify SSI recipients'
continuing eligibility for benefits. Although the agency recently obtained
"online access" to the NDNH for the DI program, this type of access only
allows SSA to obtain wage data on case-by-case basis; it does not permit
the agency to systematically match all DI beneficiaries against the NDNH
to identify those with high levels of earnings-a potentially valuable,
costeffective means of identifying beneficiaries who may be at risk for
large overpayments.

    SSA Lacks an Effective Screening Mechanism to Identify Cases Most Likely to
    Incur Large Overpayments

The agency lacks the means to identify beneficiaries who are most likely
to incur large overpayments. SSA currently uses the enforcement operation
to select individuals with more than $4,860 in annual earnings for a work
CDR. While periodic computer matches with the NDNH would help provide more
timely, comprehensive earnings data to SSA, some SSA officials told us
that the agency would still need the ability to systematically screen the
cases to identify those at high-risk for large overpayments. The agency
currently uses a screen for its medical CDR reviews, which helps the
agency identify beneficiaries who are most (or least) likely to have
medically improved.23 This screen helps SSA prioritize the use of limited
staff resources by scheduling beneficiaries who are identified as least
likely to improve for less frequent medical CDRs, and using forms that are
periodically mailed to them requesting information on their medical
condition. While our prior work has identified some problems with this
screening mechanism,24 in general SSA believes that, in many instances, it
helps the agency mitigate the need for costly, timeconsuming medical
examinations that may not be necessary. However, the agency does not
currently have a similar tool for its work CDRs to identify

22Batch file computer matches would allow SSA to periodically match all DI
beneficiaries against wage and new hires data in the NDNH, thus helping
the agency identify individuals with high levels of earnings (those who
may be likely to incur large overpayments).

23This mechanism involves the application of statistical formulas that use
data on beneficiary characteristics contained in SSA's computerized
records-such as age, impairment type, length of time on the disability
rolls, previous CDR activity and reported earnings-to predict the
likelihood of medical improvement and, therefore, benefit cessation.

24 GAO, Social Security Disability: Reviews of Beneficiaries' Disability
Status Require Continued Attention to Achieve Timeliness and
Cost-Effectiveness, GAO-03-662 (Washington, D.C.: July 24, 2003).

beneficiaries with high levels of earnings or other characteristics that
may contribute to large overpayments.

One program service center we visited is considering the use of a screen
that would give higher priority to developing cases for beneficiaries with
higher earnings, and thus the potential for larger overpayments.25 Some
SSA officials we interviewed told us that such a screen would help the
agency prioritize this workload and make better use of limited resources,
particularly in field offices where staff are often constrained by several
competing workloads, such as processing initial claims. Further, one
official in this program service center told us some of the other program
centers were considering implementing this screen.

Even if a screen existed that would allow SSA to identify cases with the
greatest potential for large overpayments, the agency still lacks a timely
alert mechanism to notify field offices and program service centers about
such cases.26 According to some SSA officials, such a mechanism, if
created, could allow the agency to quickly notify field offices and
program service centers about cases that have been identified as
high-priority for work CDRs. SSA currently uses an alert mechanism in its
SSI program to rapidly notify field offices about recipients with high
levels of earnings or other factors that may affect their eligibility for
benefits. These alerts are generated centrally from SSA's match with the
NDNH and sent electronically to field office staff, telling them which
recipients should have their cases reviewed. However, a similar alert
system does not currently exist in the DI program. Instead, SSA field
offices rely on daily workload management listings of potential work CDR
cases that are relayed via existing agency systems. These lists summarize
the cases that are awaiting review, including the "age" of the case. On
the basis of such lists obtained from several field offices, we found that
half of the cases were at least 117 days old. Moreover, cases that were
transferred from program service centers were generally older-some were
listed as being 999 days old. In addition, because the data field for
measuring the age of cases on the workload management lists only holds a
maximum of 3

25 According to SSA, the agency is currently developing a scoring
mechanism that would help the agency better manage overpayment-related
workloads in its program service centers and field offices.

26SSA's enforcement operation does generate alerts for cases that should
be reviewed. However, this alert is not as timely as the wage alerts
generated in the SSI program that notify staff about cases that should be
examined due to recipient earnings.

characters, SSA officials told us that these cases were likely older than
indicated on the lists. In addition, we found that these lists do not
allow managers to identify cases with the greatest potential for
overpayments. As a result, staff generally review cases as they are
released by managers to be developed. While some managers and staff we
interviewed told us that they make a concerted effort to review the oldest
cases first, others told us that they generally process the cases on a
"first-in-first-out" basis, which is not necessarily related to the age of
the case.

    Weaknesses in SSA's Management Information Hinder Its Ability to Effectively
    Monitor Work CDR's

Our review suggests that SSA relies on potentially inaccurate data to
manage its work CDR workload. In particular, our work shows that highlevel
management information data on the age of work CDR cases may not
accurately reflect the true age of cases (i.e. the actual time it took to
complete these cases) and may result in cases being counted more than
once, thus distorting the information that SSA relies on to measure the
number of cases that are reviewed and completed. To test the accuracy of
high-level management data for this workload, we conducted an in-depth
examination of 71 randomly selected cases27 that were "cleared" from SSA's
Processing Center Action Control System28 (the system) during a 1week
period in April 2004. On the basis of our sample we estimate that,
overall, 49 percent29 of these cases were improperly cleared from this
system. This means that the cases were listed as having been fully
reviewed and completed, when in fact they still required additional
development. Improperly cleared cases can have several negative impacts
according to SSA officials, including the potential for contributing to
large overpayments. For example, on the basis of our sample, we estimate
that 13 percent of the cases were improperly cleared from the system
because

27We worked with SSA to select a 1 percent sample of cases that were
cleared from its Processing Center Action Control System over a 1-week
period in April 2004. This 1 percent sample included a total of 151 cases.
We then randomly sampled 71 of these cases for review. For this file
review, the margin of error for all percentage estimates does not exceed
plus or minus 10 percentage points at the 95 percent confidence level,
unless otherwise noted.

28This system is used by SSA to produce management information in its
"Workload Status Report" used by the agency to track the disposition of
cases that are sent to the program service centers for development as a
result of the enforcement operation. According to SSA officials, cases
should only be listed as "cleared" in this system if they have been fully
developed and completed in accordance with agency guidelines for
processing work CDRs.

29The 95 percent confidence interval surrounding this estimate ranges from
37 to 61 percent.

they were not fully developed and did not have a "diary" attached to
them-an automated notice that reminds staff to review the case after a
specified period of time. Such cases might not be selected for review
again until the next enforcement match-which could be as much as 1 year-
and result in overpayments if the beneficiary had earnings that exceeded
the SGA level. Most importantly, because they did not have a diary, SSA
did not have any way of monitoring these cases or ensuring that they were
properly completed. This weakness may partially explain the type of old
cases with large overpayments we identified in several SSA field offices.

In addition to the cases without any diary, an estimated 37 percent30 of
the cases were incorrectly shown as cleared while still being developed in
various locations such as field offices.31 Although these cases had a
diary- thus giving SSA some level of internal control over them-our
analysis shows they could still result in management information that does
not show the true age of work CDR cases. More specifically, these types of
cases would likely result in management information that understates the
true age of such cases, and would distort the overall measurement of
progress in handling work CDR workloads. Such cases could also result in
the double counting of work CDRs. For example, if a single case was
cleared in the program center and subsequently developed and cleared in a
field office, it could be incorrectly listed as two separate work CDRs.
SSA officials also acknowledged that some of the cases we reviewed which
showed indications of being cleared multiple times could result in their
being counted as numerous separate work CDRs. Thus, existing high-level
management data may not accurately capture how many work CDRs have
actually been completed.

We also found that SSA does not currently have the capability to track the
disposition of work CDR cases. For example, the agency is unable to
systematically track how many work CDR cases involve overpayments. Because
it lacks sufficient management information data on this workload, the
agency also does not have performance goals for work CDRs similar to
measures it maintains for its medical CDR workloads, such as the number

30The 95 percent confidence interval surrounding this estimate ranges from
26 to 49 percent.

31We estimate that 28 percent of the cases were properly cleared (the 95
percent confidence interval surrounding this estimate ranges from 18 to 40
percent). In addition to the 49 percent that were improperly cleared, an
additional 14 percent had some other processing problem, while 6 percent
could not be completed on the basis of available information, and 3
percent did not have any work issue involved.

of work CDRs that should be completed each year. Moreover, given the
problems we identified with potential double or multiple counting of work
CDR cases, it is unclear whether SSA could establish meaningful
performance goals at this time.

    Identified Vulnerabilities Contribute to Some Large Overpayments

The vulnerabilities we identified have likely contributed to old work CDR
cases and large earnings-related overpayments in the DI program. We
identified several examples of cases that took years to develop and
complete. Some of these cases were as much as 7 years old and involved
large overpayments. The following are examples of some of these cases:

o  	One case we observed was initially selected for review in 1997 by the
enforcement operation. Although the recipient's benefits should have been
discontinued in 1997 according to SSA officials, payments continued until
March 2000. Agency officials could not explain why no action was taken on
this case between 1997 and 2000. As a result, this beneficiary incurred a
$28,000 overpayment.

o  	Another case was selected for review each year from 1997 to 2001 by
the enforcement operation. However, there was no evidence in the file that
a work CDR was conducted until February 2004, and SSA officials were
unable to explain why no action was taken after several consecutive
enforcement matches. This beneficiary incurred an estimated $105,000
overpayment between April 1997 and December 2003.

o  	Another case involved a beneficiary who had earnings well above SGA in
1998 when they first became eligible for DI benefits. However, the case
only arrived in the field office for action in March 2003. SSA
discontinued the recipient's benefits in 2003, but SSA officials could not
explain why the case took 5 years to arrive at the field office for
action. As a result, the recipient incurred a $32,000 overpayment.

o  	An additional case we identified involved a beneficiary with earnings
well above SGA for several years and who incurred a prior earningsrelated
overpayment. SSA subsequently waived the overpayment. However, the
recipient continued to work without reporting the earnings to SSA. The
agency eventually discontinued the individual's benefits in September
2003. At the time of our review, SSA officials estimated that the
beneficiary had incurred a $102,000 overpayment.

Further compounding the vulnerabilities that contribute to aged cases and
large overpayments, our review suggests that SSA has difficulty balancing
competing workloads. In particular, SSA field office staff are required to
perform numerous duties, including processing initial claims, serving
individuals who walk into the field office without an appointment, meeting
with beneficiaries who have requested an appointment, and processing the
"special disability workload." Many managers and staff we interviewed told
us that work CDRs generally receive lower priority than some of these
other activities, such as processing initial claims. In several offices we
visited, we observed lists of pending work CDRs, sometimes stored in file
cabinets for extended periods of time.

SSA is currently implementing a new automated system that may address some
of the vulnerabilities we identified. This system, called "eWork," is
intended to simplify how SSA manages and processes its disability cases.
In particular, according to documentation provided by SSA, this system
will establish program controls for all work CDR cases and help the agency
identify higher priority cases. Once fully implemented, eWork will combine
data from several different SSA databases and will automate the processing
of numerous forms commonly used in developing and documenting disability
cases, according to SSA. One field office we visited was piloting this
system. Management and staff in this office generally reported that the
system was an improvement over existing systems. In particular, officials
reported that the system was useful in helping them track the age of work
CDR cases, especially older cases that should potentially receive higher
priority. Overall, SSA management and line staff expressed confidence that
this new system will improve the agency's ability manage its disability
cases, including work CDRs. However, because the system is new and is not
yet fully implemented nationwide, we were unable to evaluate how effective
it may be for addressing some of the weaknesses we identified.

Conclusions 	We recognize that ensuring program integrity while focusing
on the important goal of returning individuals with disabilities to work
presents challenges for SSA. However, the weaknesses we identified in
SSA's existing work CDR processes continue to expose the program to
overpayments and abuse. In particular, SSA's reliance on outdated earnings
information has contributed to overpayments and forced staff to
investigate cases that are old and thus difficult and time-consuming to
process. Without the ability to conduct batch file computer matches with

the National Directory of New Hires, the agency will remain vulnerable to
large earnings-related overpayments. Similarly, the lack of a screen to
systematically identify beneficiaries more likely to incur overpayments
means that SSA cannot target cases that should receive higher priority.
Even if such a screen existed, SSA would not be able to make the best use
of it given the lack of an automated alert system to notify field offices
and program service centers about which cases should be reviewed.
Moreover, without accurate, reliable management data on the age and status
of work CDR cases, SSA will find it difficult to effectively monitor this
workload, identify areas that require continued improvement, and develop
meaningful work performance measures.

In an environment of limited budgetary and staff resources, federal
agencies such as SSA will be required to take a more strategic approach to
servicing ever-increasing workloads. The magnitude of earnings-related
overpayments indicates that SSA should take additional steps to strengthen
DI program integrity. Moreover, the potential of having to repay a large
overpayment may discourage some beneficiaries from continuing to work,
thus working contrary to SSA's goal of helping individuals become
self-sufficient. Ultimately, without a concerted effort to increase
management focus on this key workload and to reengineer existing
processes, SSA's ability to ensure that trust fund dollars are protected
and reserved for those who are truly eligible will continue to be
compromised. The new automated system that SSA is developing may help the
agency address some of the weaknesses we identified, but it is too early
to determine how effective it will be. A conscious management decision to
use this system to improve DI program integrity in conjunction with more
accurate management information will be required to help detect and
prevent large overpayments.

To enhance SSA's ability to detect and prevent overpayments in the DI
program, we recommend that the Commissioner of Social Security take the
following actions to improve the agency's work CDR processes:

1. 	Initiate action to develop a data sharing agreement with the Office of
Child Support Enforcement to conduct batch-file periodic computer matches
with the National Directory of New Hires (NDNH). Such matches would
provide SSA with more timely data to help the agency systematically
identify DI beneficiaries who are most likely to incur overpayments. Such
a tool could also allow SSA to perform a one-time, comprehensive match
against all DI beneficiary records to identify individuals who may be
overpaid but have not yet been detected.

  Recommendations

  Agency Comments
  and Our Evaluation

2. 	Consider developing an enhanced screening mechanism that would enable
the agency to more effectively identify DI beneficiaries who are most
likely to incur earnings-related overpayments. This would help the agency
make more efficient use of limited staff and budgetary resources.

3. 	Study the potential for creating an alert system similar to that used
in the SSI program for alerting field offices about recipients at high
risk for earnings-related overpayments. Such a system would allow SSA to
notify field offices and program service centers about beneficiaries the
agency identifies as most likely to incur large overpayments.

4. 	Consider ways to improve the accuracy and usefulness of existing
management information data. Improvements may include modifying how the
agency measures the age of work CDR cases to more accurately reflect how
long they are in process.

5. 	Once the eWork system is fully implemented, SSA should consider how it
could be used to help the agency create performance goals for its work CDR
workload.

We provided a draft of this report to SSA for review and comment. SSA
agreed with our recommendations and, in some instances, outlined initial
plans for their implementation.

SSA agreed with our first recommendation to develop a data-sharing
agreement with the Office of Child Support Enforcement to conduct
batch-file computer matches with the NDNH. The agency noted that it
pursued online access to the NDNH first because it was more cost effective
and expeditious. The agency also indicated that it is developing a new
computer matching agreement that supports SSA's use of the NDNH in the DI
program for purposes of identifying potential overpayments. We encourage
SSA to ensure that any new agreement will provide for periodic, batch-file
matches to verify beneficiaries' earnings at regular, specified intervals.

SSA also agreed with our second recommendation to consider developing an
enhanced screening mechanism to help the agency more effectively identify
DI beneficiaries most likely to incur earnings-related overpayments. In
particular, SSA agreed that it should pursue a screening system similar to
that currently used for medical CDRs to determine if there is an increased
likelihood of earnings-related overpayments based on particular diagnosis
codes. It also noted that it should study ways to

improve the effectiveness of existing systems (such as the Continuing
Disability Review Enforcement Operation and the Disability Control File)
to help the agency focus on beneficiaries with the greatest potential for
overpayments. We agree that these are positive steps and that the agency
should consider how improvements to such systems might be incorporated to
emerging systems such as "eWork".

With respect to our third recommendation that SSA develop an alert system
similar to that currently used in the SSI program for alerting staff to
cases at risk for earnings-related overpayments, SSA agreed and noted that
an alert system such as the "S2" alert used for SSI wage discrepancies
could provide a useful model. The agency noted that such an alert could
reduce the amount of time in which a claimant would continue to receive
payments while work development is initiated. Moreover, since these
reports include the employer's name, address, and a quarterly breakdown of
the beneficiary's earnings, this detailed information would provide SSA
staff with more specific information than is currently available. SSA also
said that an alert system could also be generated from the NDNH match
proposed in our first recommendation. We agree that an alert system would
help identify potential overpayments more quickly, particularly if it were
generated from data produced by periodic computer matches with the NDNH.

SSA agreed with our fourth recommendation to improve the accuracy and
usefulness of existing management information data. SSA said that it is
working on a plan to unify the manner in which it identifies and counts
work and medical CDRs. The agency believes that it will be able to more
accurately capture workload counts and employee time consistently,
regardless of where the work is performed. While we agree that efforts to
improve existing processes and systems are necessary, it is too early to
determine if the proposed modifications will address the problems we
identified with high level management information, such as potential
double-counting of work CDRs.

SSA also agreed with our fifth recommendation to consider how it could use
the eWork system to create performance goals for work CDRs once it is
fully implemented. The agency commented that such a measure would give
field offices and program service centers a better indication of what is
expected of them regarding processing this workload and would help them
balance the time needed to process competing workloads.

SSA's formal comments appear in appendix II. SSA also provided additional
technical comments that we have incorporated in the report as appropriate.

Unless you publicly announce its contents earlier, we plan no further
distribution until 30 days after the date of this report. At that time, we
will
send copies of this report to the House and Senate Committees with
oversight responsibility for the Social Security Administration. We will
also make copies available to other parties upon request. In addition, the
report will be available at no charge on GAO's Website at
http//:www.gao.gov. If you have any questions concerning this report,
please contact me at (202) 512-7215.

Sincerely yours,

Robert E. Robertson
Director, Education, Workforce,

and Income Security Issues

                       Appendix I: Scope and Methodology

This appendix provides additional details about our analysis of the
Disability Insurance (DI) program's work continuing disability review
(work CDR) process, including potential weaknesses in the Social Security
Administration's (SSA) existing procedures and policies. To meet the
objectives of the review, we examined DI performance data, prior reports
by SSA and its Office of Inspector General (OIG), external research
studies, and our prior reviews of the program. We analyzed DI payment data
over a 5-year period from 1999 to 2003, and examined cases from 14 out of
18 SSA field offices we visited and 3 program service centers. In
addition, we randomly selected and reviewed 71 cases with earnings to
determine if they were reviewed and processed in accordance with program
guidelines. Finally, we conducted in-depth interviews with 230 management
and line staff from SSA's headquarters; its regional offices in New York
and San Francisco; 18 field offices in 6 states; and 3 out of 8 regional
program service centers. During our meetings, we (1) examined existing
work CDR procedures; (2) documented management and staff views on the
effectiveness of SSA's work CDR processes for detecting and preventing
earnings-related overpayments; and (3) discussed potential improvements to
existing program processes, systems, and policies.

We conducted independent audit work in six states (California, Florida,
Maryland, Massachusetts, New York, and Virginia) to examine SSA's policies
and procedures for conducting work CDRs, and to identify any common
weaknesses in SSA's work CDR processes. We selected locations for field
visits based on several criteria, including geographic dispersion, states
with an SSA program service center, states with large numbers of DI
beneficiaries, and states with large DI expenditures. In total, we visited
18 field offices and interviewed 161 SSA field office managers and line
staff responsible for the DI program. We visited a mix of large offices in
metropolitan areas as well as smaller offices located in the suburbs. In
addition, we visited three program service centers in Richmond,
California; Queens, New York; and Baltimore, Maryland. These program
centers were responsible for the majority of all work CDRs identified by
the enforcement operation. Where appropriate, we also visited field
offices or program centers that were conducting special initiatives or
piloting emerging computer systems that could impact how SSA conducts work
CDRs (such as the "eWork" system).

During our meetings with SSA and OIG officials, we documented management
and staff views on the effectiveness of work CDR policies and procedures
and potential improvements to existing processes, policies, and systems.
In particular, we documented management and staff views on (1) the
timeliness of existing data sources to verify beneficiary

Appendix I: Scope and Methodology

earnings, (2) the effectiveness of existing processes for identifying
individuals at high-risk for large overpayments, (3) the effectiveness of
existing computer systems for notifying staff responsible for conducting
work CDRs about cases that should be reviewed, and (4) the accuracy of
management information data used to monitor work CDRs in one large program
service center. To further assess existing program processes and systems,
at 10 offices and 3 service centers, we judgmentally selected between 5
and 7 pending or completed work CDR cases. We generally looked at older
cases in order to understand where existing procedures may have
weaknesses. We then conducted in-depth reviews of these case files to
identify potential vulnerabilities in existing work CDR processes,
policies, and systems.

As part of our study, we worked with SSA to draw a 1 percent sample of all
work CDR cases that were "cleared" from the agency's Processing Center
Action and Control System over a 1-week period in April 2004 (the "study
population"). Our objective was to determine whether work CDR cases were
cleared in accordance with agency guidelines and to assess the accuracy of
high-level management data produced by this system. This sample resulted
in a total of 151 cleared cases. We then randomly selected 71 of these 151
cases for review. As part of our review, we discovered that there was a
potential for cleared work CDR cases to appear multiple times in SSA's
Processing Center Action Control System. On the basis of our discussion
with knowledgeable SSA officials, we determined it would be highly
unlikely for cases to be listed as "cleared" multiple times in a 1week
time period. Therefore, we assumed that cases did not appear more than
once in the 1-week time period from which we drew our sample.

Because we followed a probability procedure based on random selections,
our sample is only one of a large number of samples that could have been
drawn. Since each sample could have provided different estimates, we
express our confidence in the precision of our particular sample's results
using 95 percent confidence intervals. A confidence interval is an
interval that would contain the actual population value for 95 percent of
the samples we could have drawn. As a result, we are 95 percent confident
that each of the confidence intervals in this report will include the true
values in the study population. For this file review, the margin of error
for each percentage estimate does not exceed plus or minus 10 percentage
points, unless otherwise noted. The margin of error is the distance from
each estimate to the upper or lower boundaries of its 95 percent
confidence interval.

Appendix I: Scope and Methodology

To assess the reliability of the databases we used, reviewed reports
provided by SSA and its Office of Inspector General, which contained
recent assessments of these databases. We also interviewed knowledgeable
agency officials to further document the reliability of these systems. In
addition, we checked the data for internal logic, consistency, and
reasonableness. We determined that all the databases were sufficiently
reliable for purposes of our review.

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix II: Comments from the Social Security Administration

Appendix III: GAO Contacts and Staff Acknowledgments

GAO Contacts 	Daniel Bertoni, Assistant Director (202) 512-5988 Jeremy D.
Cox, Analyst-in-Charge (202) 512-5717

Staff In addition to those named above, Jeff Bernstein, Sue Bernstein, Dan
Schwimer, Salvatore F. Sorbello, Sidney Schwartz, and Shana Wallace
Acknowledgments made important contributions to this report.

Related GAO Products

Social Security Disability: Reviews of Beneficiaries' Disability Status
Require Continued Attention to Achieve Timeliness and Cost-Effectiveness.
GAO-03-662. Washington, D.C.: July 24, 2003.

High-Risk Series: An Update. GAO-03-119. Washington, D.C.: January 2003.

SSA Disability: Enhanced Procedures and Guidance Could Improve Service and
Reduce Overpayments to Concurrent Beneficiaries. GAO-02802. Washington,
D.C.: September 5, 2002.

Social Security Administration: Agency Must Position Itself Now to Meet
Profound Challenges. GAO-02-289T. Washington, D.C.: May 2, 2002.

Social Security Disability: Disappointing Results from SSA's Efforts to
Improve the Disability Claims Process Warrant Immediate Attention.

GAO-02-322. Washington, D.C.: February 27, 2002.

Social Security Administration: Status of Achieving Key Outcomes and
Addressing Major Management Challenges. GAO-01-778.
Washington, D.C.: June 15, 2001.

High-Risk Series: An Update. GAO-01-263. Washington, D.C.: January 2001.

Major Management Challenges and Program Risks: Social Security
Administration. GAO-01-261. Washington, D.C.: January 2001.

Social Security: Review of Disability Representatives. GAO/HEHS-9950R.
Washington, D.C.: March 4, 1999.

Major Management Challenges and Program Risks: Social Security
Administration. GAO/OCG-99-20. Washington, D.C.: January 1999.

High-Risk Program: Information on Selected High-Risk Areas. GAO/HR97-30.
Washington, D.C.: May 16, 1997.

High-Risk Series: An Overview. GAO/HR-97-1. Washington, D.C.: February
1997.

  GAO's Mission

Obtaining Copies of GAO Reports and Testimony

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site (www.gao.gov). Each weekday, GAO posts newly
released reports, testimony, and correspondence on its Web site. To have
GAO e-mail you a list of newly posted products every afternoon, go to
www.gao.gov and select "Subscribe to Updates."

                             Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: 	Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

  To Report Fraud, Contact:

Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470

Gloria Jarmon, Managing Director, [email protected](202)
512-4400Congressional U.S. Government Accountability Office, 441 G Street
NW, Room 7125 Relations Washington, D.C. 20548

Public Affairs 	Jeff Nelligan, Managing Director, [email protected] (202)
512-4800 U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548

                             Washington, D.C. 20548
*** End of document. ***