The purpose of this study was to determine the number and characteristics
of suspended or terminated Supplemental Security Income (SSI) records
having an unresolved Internal Revenue Service (IRS) alert indicator.
The SSI program is administered by the Social Security Administration
(SSA) and provides cash assistance to people who are aged, blind,
or disabled whose income and resources are limited. Since SSI eligibility
is based on need, the SSA must be aware of any changes in recipients` income
or resources. The Tax Reform Act of 1984 provides for SSA to receive
financial information from the IRS to help detect unreported nonwage
information, such as pensions, interest, and dividends. When IRS
records indicate possible income or resources not reported on their
SSI records, an alert is generated (referred to as a 5B alert) and
the recipient`s SSI record is annotated. The SSA field offices
(FO) are required to verify the information with the recipient or
the financial institution and adjust or terminate the benefits if
warranted. If the recipient fails to cooperate with SSA, benefits
are suspended and eventually terminated without determining whether
an overpayment exists for the earlier period.
From a 5 percent sample of all SSI records, we requested those pertaining
to recipients with a suspended or terminated record containing a
5B alert. We obtained and reviewed financial records for these individuals
and analyzed the information for possible overpayments and potential
fraud. We also surveyed a sample of SSA FOs to obtain their perspectives
of the 5B alert process.
In 1992, There Were Approximately 143,000 Recipients That Had
A Terminated SSI Record With An Unresolved 5B Alert.
There were 7,152 recipients in our sample (which projects to 143,040
5B alerts, 2.6 percent of the SSI population) that were not fully
developed or resolved before payments were stopped. Even though SSA
has since revised its alert selection procedures and enhanced the
efficiency of the 5B alert program, we believe that the 5B alert
process remains vulnerable to fraud and abuse due to SSA`s limited
authority to investigate IRS alert cases.
Seventeen Of The 42 SSI Recipients For Which We Obtained Income
And Resource Information May Have Received Overpayments.
According to SSA, an average overpayment for an SSI recipient with
a 5B alert is $1,094. The potential overpayments in the cases we
reviewed were $2,178.
SSAs FO Staff Are Not Following Program Operations Manual
System (POMS) Procedures In Some Cases Or Fully Utilizing Resources
Available To Them For Processing 5B Alerts.
One-fourth of the FOs do not fully develop 5B alerts. Two-thirds
of the FOs do not impose financial penalties for recipients who do
not report all resources. Also, variations exist in applying administrative
finality which restricts SSA to 24 months after the SSI record is
terminated to take corrective action on an overpaid case, except
where fraud or similar fault exists.
One-Third Of The Sampled Recipients With 5B Alerts Also Receive
Social Security Benefits, A Possible Avenue To Recovering SSI Overpayments.
Recovery of overpaid title XVI funds can be made from Social Security
benefits, but currently only when beneficiaries request this means
FOs Identified Barriers That Impede The Development Of 5B Alerts.
FO staff identified many barriers that inhibit timely and complete
development of 5B alerts, including difficulties in obtaining information
from financial institutions and a lack of staff to process them.
SSA has taken steps to enhance the efficiency of the 5B alert
program. This change will reduce the number of alerts by up to
75 percent. However, we believe there are additional steps that
SSA can take to improve program effectiveness.
consider requiring FOs to develop previous 5B alerts to
the extent possible prior to allowing recipients to be re-entitled
encourage increased utilization of financial penalties for
those recipients who receive overpayments detected by 5B alerts;
provide refresher training for FO staff to ensure compliance
with SSA procedures for developing 5B alerts;
determine whether work unit credits for developing 5B alerts adequately
reflect the effort extended by FO staff;
continue to work towards legislative changes permitting
nonvoluntary recovery of overpayments by adjusting Social Security
pursue legislative changes in the SSI program to simplify
resource and income eligibility requirements;
build on SSA`s working relationship with national banking
associations to improve cooperation and support for this workload
by financial institutions; and
refer all suspected fraud cases to the Office of the Inspector
SSA concurs with the intent of our recommendations to maximize the
intended benefit from procedures related to overpayments detectable
by IRS alerts. The Agency expressed concerns about the methodology
we used and some of the contents of the report. We have addressed
those concerns in the Recommendations section of the report.
The purpose of this study was to determine the number and characteristics
of suspended or terminated SSI records having an unresolved IRS alert
The SSI program is authorized by title XVI of the Social Security
Act (SSI for the aged, blind, and disabled) under legislation passed
in 1972. The SSI provides cash assistance to people who are aged,
blind, or disabled and whose income and resources are limited.
Because SSI benefits are needs-based, SSA, which administers the
SSI program, must be aware of changes in recipients` income or
resources. The Tax Reform Act of 1984 provides for SSA to receive
financial information on recipients` nonwage earnings (pensions,
unemployment compensation, interest, dividends, bank accounts, trusts,
or stocks) through computer interfaces with IRS. Three times a year,
SSA performs a match of SSI recipients against the most recent IRS
data available to determine if there is a need to adjust or terminate
payments. The law permits SSA to suspend payments to a recipient
who does not respond to requests for it`s review of their financial
Developing a 5B Alert
When IRS data indicates SSI recipients have income or resources
not reported on their SSI records, this generates what is referred
to as a 5B alert. This alert produces a diary that is maintained
on the SSI record. The diary alerts the field staff that they must
verify with the recipient the unreported source of income or resources
and may need to adjust or terminate the recipient`s benefits.
The servicing FO contacts the SSI recipient to request permission
to review his or her monthly financial income records to verify the
source of income.
There are instances when a 5B alert may not indicate an overpayment.
1. resources below program limits, 2. an increase
due to burial funds which are excluded from the definition of resources,
3. a recipient can prove that funds in a joint account belong
to another party, 4. funds are part of a plan for achieving self-support
which are excluded from an eligibility determination, or 5. the
case was in nonpay status during the alert period in question.
To determine if there is an overpayment, a recipient`s financial
records must be reviewed. The Right to Financial Privacy Act of 1978,
Public Law 95-630, restricts SSA`s access to financial records
unless the recipient provides authorization voluntarily. If the recipient
cooperates, the case is developed and, if overpayments are identified,
payment adjustments are made. If the recipient does not respond or
refuses to cooperate, SSA is required to give 30 days notice prior
to taking adverse action. If there is no response to it`s initial
request and follow-up, SSA is required to suspend payments beginning
with the next month or the following month for the recipient failing
to provide evidence of continuing eligibility.
In cases where recipients do not provide voluntary consent to the
review of their financial records, there are four other mechanisms
through which the Government may obtain access to financial records:
1. administrative summons or subpoena, 2. search warrant, 3. judicial
subpoena, and 4. a formal written request.
The majority of these
mechanisms are not available to all Government agencies. Administrative
requests are commonly used in connection with legitimate law
enforcement inquiries. OIG may access financial records through administrative
subpoena. OIG`s Office of Investigations (OI) may also utilize
search warrants and judicial subpoenas with the Department of Justice
approval. Although SSA Regional Commissioners have administrative
subpoena power, it is rarely used. SSA FOs can only use customer
When a recipient has failed to cooperate and payments have been
suspended for 12 months, the record is terminated. If the recipient
files a new application after the record is terminated, there is
no requirement for SSA to investigate or resolve the prior alert.
Twenty-four months after the 5B alert is terminated, the record is
barred from being corrected due to administrative finality except
in cases where fraud or similar fault (knowingly concealing information
that is material to an SSA determination) can be established.
An SSA FO can make similar fault determinations and refer cases
to OI when it appears that fraud is involved. Fraud referral guidelines
are found in SSA POMS.
We obtained and reviewed sections of POMS that pertain to IRS alerts
and interviewed SSA regional office personnel that are involved in
the implementation and interpretation of SSA policy. We surveyed
a sample of FOs to obtain their perspectives of the 5B alert process.
We also visited an SSA FO where we interviewed SSA FO personnel who
are directly involved in SSI and handle IRS alerts.
From one segment (a 5 percent sample) of the SSI population, we
requested those records that had been terminated or suspended with
5B alerts along with any other records for those individuals. There
were 7,093 recipients in this sample segment that had a single 5B
alert while 59 had multiple alerts which is roughly 2.57 and .03
percent, respectively, of the general SSI population. From the sample,
we selected two sub-samples of recipients, one with single alerts
and the other with multiple alerts. Table 1 explains the breakdown
of our sub-samples and the cases we were able to actually review
as part of our inspection.
BREAKOUT OF IRS ALERT SUB-SAMPLES
Original number in sub-sample
Less cases previously developed, deceased,
age 75 or older, or insufficient information in file
Less case files not in terminated or suspended
Case files reviewed in our inspection
For each recipient with either single or multiple 5B alerts, we
obtained case files and financial records when possible. We also
obtained a SSI Record Display (SSIRD) for each recipient, which contains
information on payment status, denial codes, resources, eligibility
date, earned income, and overpayment/underpayment amounts.
We contacted the recipients in our sample to request permission
to obtain information regarding resources and income from their financial
institution. We submitted administrative subpoenas to those financial
institutions when a recipient did not provide authorization. Further,
we submitted administrative subpoenas to "other entities" such
as insurance companies or trading companies, requesting financial
SSA FO Survey
We surveyed a sample of FOs with 5B alerts to obtain their perspectives
of the alert process. From a universe of 638, we randomly selected
a sample of 105 FOs. We requested the name of an individual in each
office who was knowledgeable in the processing of 5B alerts. We mailed
a survey instrument to each individual and then analyzed the responses.
The 5 percent sample of the SSI population was selected during 1992.
The FO survey was conducted during 1993. The SSIRDs for the alert
cases were obtained during 1993. The case files for the alert cases
were reviewed during 1994. Our review was conducted in accordance
with the Quality Standards for Inspections issued by
the President`s Council on Integrity and Efficiency.
In our sample, there were 7,152 recipients which projects to 143,040
5B alerts that were not fully developed or cleared by SSA FOs, or
2.6 percent of the SSI population. As a result of our review and
analysis, we found that the 5B alert process is vulnerable to abuse
and fraud due to SSA`s limited authority to investigate 5B alerts.
Our analysis of the SSI records and financial information revealed
some possible overpayments. We were not able to conclusively establish
that overpayments existed in these cases because we did not contact
the recipients to determine ownership of the income/resources.
One recipient received a check for $25,120.82, 6 months
prior to his or her date of entitlement. The recipient withdrew
$8,816, 2 months prior to the date of entitlement. Deposits amounting
to $14,738 were made to the account after SSI payments were terminated.
Another recipient had over $9,000 in savings accounts which
was withdrawn the month prior to the date of entitlement.
A deposit of $83,793 was made to a recipient`s account
3 months after SSI payments were terminated.
Some reasons the SSI payments in our sample were terminated include:
excess resources or income (22 percent and 27 percent respectively);
failure to provide required reports to resolve the alert (24 percent);
and death of the recipient (9 percent). The cases above are being
reviewed for fraud and overpayment recovery and may be referred to
The resource limitation for individuals on SSI is a $2,000 average
monthly account balance. According to SSA, an average overpayment
for an SSI recipient with a 5B alert is $1,094. We found in our inspection
that the average possible overpayment for those recipients with average
monthly account balances over $2,000 was $2,178. This involved both
recipients with single and multiple unresolved IRS alerts.
There currently is no requirement that the FO develop an unresolved
5B alert on a prior record when a recipient comes in to reapply for
SSI. Although not required, the majority of the FOs surveyed do check
the previous record and attempt to develop such alerts. Table 2 details
single and multiple unresolved IRS alert cases in our sample that
may have incurred overpayments. For our analysis, overpayments were
estimated by reviewing financial records for months where recipients` received
SSI payments when their account balances were over $2,000. The average
SSI payment was multiplied by the number of months when their account
balance exceeded $2,000 to estimate the overpayment.
Unresolved 5B Alerts With Estimated
No. of months with balance over $2,000
Avg. monthly account balance
Avg. Monthly account balance
No. of months in pay status
1 1 2
5 1 2
1 Involves a minor child.
2 Overpayment may exceed estimate - incomplete
We found that some FOs are not following POMS procedures in
documenting, clearing, and retaining 5B alerts.
Twenty-five FOs (24 percent) do not document the case files
with reasons for not fully developing an alert, nor do they clear
these cases from the system when appropriate.
Seventy-nine (75 percent) of the 105 FOs surveyed continued
to retain the 5B alert information in the file for developed cases.
According to the Computer Matching and Privacy Protection Act of
1988, Public Law 100-503, the FO is required to destroy the 5B
alert upon completion of development.
The FOs are not fully utilizing financial penalties and similar
Sixty-nine (66 percent) of the FOs surveyed do not impose
penalty deductions, while 28 (27 percent)) impose these 10 percent
of the time or less.
Fifty-eight (55 percent) never make a determination of
similar fault and 31 (30 percent) make such a determination in
10 percent or less of the time in their cases.
We found variation among FOs in terms of how they describe
and apply administrative finality rules.
The claims representatives go back 26 months from
the date on the actual alert. If that is not available, they use
the 5B diary date on the SSIRD."
"We use 25 months prior to the run-date selection
of the 5B alert."
"Go back 23 months from date the diary is set, per
POMS, unless fraud is involved."
Recovery of overpaid title XVI funds can be made from title II benefits
if the recipient agrees to make repayment by decreasing these benefits.
Under current legislation, recovery in this manner can only be utilized
in cases when beneficiaries volunteer to offset their Social Security
Staff in 20 of the surveyed FOs (19 percent) stated that
financial institutions were resistant to responding to information
requests. Another 13 FOs (12 percent) responded that they received
mixed responses from financial institutions. Other comments about
information requests from financial institutions include: being
slow to respond, incomplete, old and incorrect information, and
costly to obtain.
Staff in 88 of the FOs (84 percent) stated that financial
institutions indicated many of the account numbers that appear
on the 5B alert are invalid.
Eighty-six of the FOs (82 percent) have been unable to
obtain information from financial institutions in some cases. The
primary reasons given for not providing information were that the
account number was invalid, requested information was too old,
too time-consuming, bank records were unavailable, mergers made
locating information difficult, amount of fee reimbursement too
low, and they refused to respond in spite of the fee SSA was willing
Financial institutions are required to report interest
income monthly so it can be recorded on the recipient`s SSI
record. Some amounts are as low as 11 cents per month. Resources
in excess of $61 must be reported. The FOs responded this becomes
a burden on financial institutions to report these small amounts.
There are instances in which FOs do not attempt to contact
the financial institution listed on the 5B alert because they know
that it is chronically uncooperative.
Other problems include:
Seventeen FOs (16 percent) responded that recipients do
Sixty-three of the FOs (60 percent) listed problems such
as lack of staff and time to work the alerts, time-consuming, tolerances
are too low, duplicate postings occur, and they do not receive
results in relation to time spent.
Approximately one-third (39 percent) of the FOs surveyed
do not receive training for processing 5B alerts.
Seventy-seven FOs (73 percent) stated the process was not
effective in preventing overpayments.
Although IRS alert information is usually accurate, errors
do occur. For instance, bank accounts, Social Security numbers,
and recipient names do not always match. The FOs noted that double
postings sometimes occur.
The FOs do not receive work unit measurement credits for
developing 5B alerts.
Problems with the fraud referral of cases include:
Eighty of the FOs (76 percent) surveyed are not making
fraud referrals to OIG. Another 22 (21 percent) make these less
than 10 percent of the time. Many have stopped making fraud referrals
because they claim that little can be done with them and they believe
it is a waste of time. As a result, staff in 73 of the FOs (70 percent)
stated that there is little or no potential for additional fraud
Many cases are not pursued because the overpayment amounts
are below the threshold set by many U.S. Attorneys. There is no
provision for referring cases where the overpayment amount is unknown,
such as noncooperation in developing the 5B alerts. Further, many
cases are not pursued because the recipients are elderly or disabled
and their cases are difficult to prosecute.
While SSA has taken steps to enhance the efficiency of the
5B alert program, we believe there are additional steps that
SSA can take to improve program effectiveness.
1. Require FOs to develop previous 5B alerts prior to allowing
recipients to be re-entitled to SSI.
Currently there is no requirement that FOs resolve a 5B alert
on a previous record when a recipient files a new application.
At a minimum, the applicant should provide an explanation of the
2. Encourage increased utilization of financial penalties
for those recipients who receive overpayments detected by 5B alerts.
The FOs should consider increased efforts to: (a) impose penalty
deductions and make similar fault determinations and (b) place
cases in suspense status to a much greater degree than they are
now for recipients who receive overpayments based on FO development
of a 5B alert.
3. Provide refresher training for FO staff to ensure compliance
with POMS on working 5B alerts and resolving them in a timely
and thorough manner.
Since one-third of the FOs indicate they do not receive training
on 5B alerts, and one-fourth do not fully develop these alerts,
refresher training could address proper development of 5B alerts
and provide a greater awareness of the legal requirements of computer
matching and administrative finality.
4. Consider establishing work unit credits for FO staff
developing 5B alerts.
Recognizing 5B workload activities by FO staff could be accomplished
by establishing and reporting work unit measurement credits.
5. Work towards recovering overpayments by adjusting Social
Security benefits when possible.
Since one-third of the recipients with 5B alerts receive Social
Security benefits, SSA FO staff should use this method of offset,
where possible, to recover SSI overpayments.
6. Pursue legislative changes to simplify both resource
and income reporting requirements.
Simplifying or increasing the required reporting amount on income
and resources will reduce the burden on financial institutions,
as well as FO staff.
7. Establish a working relationship with national banking
associations to ensure cooperation and support on the part of the
This would facilitate and simplify obtaining information from
financial institutions in order to process the 5B alerts. In addition,
it would be beneficial in locating institutions if an on-line address
directory for financial institutions is accessible that would be
8. Refer all cases of suspected fraud to OIG.
AGENCY COMMENTS AND OIG RESPONSE
SSA comments to our draft report are contained in their entirety
at Appendix A. SSA is in agreement with recommendations three through
eight and has identified action either underway or planned to implement
The Agency points out that increased matching tolerances and a profiling
system have significantly reduced the 5B alert workload and would
significantly reduce the number of alerts (143,000) we identified
in our study. We agree.
The Agency also questions whether there were overpayments in the
17 cases we identified, pointing out that the resources may not be
owned by the recipient or may not be countable. In these cases we
went directly to the financial institutions involved to obtain income
and resource information. We requested each recipient to sign a consent
form giving permission to the financial institution to release the
information on their income and resources.
There is also concern that our finding that FOs are not clearing
5B alerts is not supported by any specific data in the report. Our
conclusions are based on a random sample of district and branch offices.
Each office was requested to have a knowledgeable staff person respond
to our survey questions. We based our conclusions on the responses
from these individuals to a wide range of questions concerning SSA
operations relative to 5B alerts.
We recommended that SSA, "Require FOs to develop previous
5B alerts to the extent possible prior to allowing recipients to
be reentitled to SSI." (Recommendation 1) We continue
to believe this recommendation is in accordance with SSAs
plans to concentrate on the most productive alerts.
SSA is concerned that our recommendation to, "Encourage
increased utilization of financial penalties for those recipients
who incur overpayments detected by 5B alerts"(Recommendation
2) infers that FOs should be encouraged to apply more penalties
for 5B cases than others. It is not our intent that FOs apply more
financial penalties for 5B cases. We are only recommending that
penalties be applied when appropriate. The staff of surveyed FOs
told us that they are not applying financial penalties.
This evaluation report was prepared by the Office of Audit under
the direction of Scott Patterson, Director, Evaluations and Technical
Services. Project staff included Brenda Stup, Senior Evaluator.
Special thanks to the following individuals from the Department of
Health and Human Services/Office of Inspector General, for their
contributions through the formal draft stage of this project:
Kansas City Region
James H. Wolf, Regional Inspector General
Jennifer E. Collins, Project Leader
Perry A. Seaton, Program Analyst
Hugh R. Owens, Program Analyst
Barbara R. Tedesco, Mathematical Statistician
Linda M. Moscoe, Program Analyst