Social Security
Better Coordination among Federal Agencies Could Reduce Unidentified Earnings Reports Gao ID: GAO-05-154 February 4, 2005Each year, the Social Security Administration (SSA) receives millions of employer-submitted earnings reports (Form W-2s) that it is unable to place in an individual Social Security record. If the Social Security number (SSN) and name on a W-2 do not match SSA's records, the W-2 is retained in the Earnings Suspense File (ESF). SSA's ability to match earnings reports is essential to calculating Social Security benefits. Because of concerns about the size of the ESF, GAO was asked to determine (1) how SSA processes workers' earnings reports, (2) the types of errors in ESF reports and the characteristics of employers whose reports are in the ESF, (3) how often earnings from repeatedly used SSNs have been reinstated and who receives the earnings from theses reports, and (4) what key factors contribute to ESF postings.
Upon receiving over 250 million earnings reports annually from employers, SSA uses various processes to post such reports to workers' Social Security records. For reports in which worker names and SSNs exactly match SSA's information, the earnings are credited to the appropriate Social Security record. When SSA encounters earnings reports that do not match its records, SSA attempts to make a match through various automated processes. Such processes have allowed SSA to identify valid records for an average of 15 million reports annually. However, about 4 percent of the reports still remain unmatched and are retained in the ESF. SSA uses additional automated and manual processes to continue to identify valid records. The most recent data show that SSA posted ("reinstated") over 2 million earnings reports in the ESF to valid records from such processes. Earnings reports in the ESF have serious data problems and are particularly likely to be submitted by certain categories of employers. Such problems include missing SSNs and employer use of the same SSN for more than one worker in the same tax year. Additional problems include missing surnames or names that include nonalphabetic characters. Forty-three percent of employers associated with earnings reports in the ESF are from only 5 of the 83 broad industry categories. Among these industry categories, a small portion of employers account for a disproportionate number of ESF reports. SSA has reinstated a substantial number of earnings reports with SSNs that appear repeatedly in the ESF. We analyzed the most frequently occurring 295 SSNs, which appeared in ESF 1,000 times or more between tax years 1985 and 2000. Of the earnings reports associated with these SSNs, SSA reinstated 13.1 million to the records of about 11.7 million workers. Although most reinstatements were for U.S.-born workers, in recent years the percentage of reinstatements to foreign-born workers has markedly increased. Also increasing is the percentage of foreign-born workers that received reinstatements for earnings in years prior to receiving a valid SSN--a potential indicator of unauthorized employment. Three major factors contribute to ESF postings. Under IRS regulations, employers must ask new hires to provide their name and SSN, but are not required to independently corroborate this information with SSA. DHS requires employers to visually inspect new workers' identity and work authorization documents, but employers do not have to verify these documents, and they can be easily counterfeited. Further, IRS regulations are minimal; IRS has no record of assessing a penalty for filing inaccurate earnings reports; and DHS enforcement efforts against employers who knowingly hire unauthorized workers has been limited in recent years because of shifting priorities following the events of September 11, 2001. Last, although SSA and DHS offer employers verification free of charge, these services are voluntary, have some limitations, and remain underutilized.
RecommendationsOur recommendations from this work are listed below with a Contact for more information. Status will change from "In process" to "Open," "Closed - implemented," or "Closed - not implemented" based on our follow up work.
Director: Team: Phone:GAO-05-154, Social Security: Better Coordination among Federal Agencies Could Reduce Unidentified Earnings Reports
This is the accessible text file for GAO report number GAO-05-154
entitled 'Social Security: Better Coordination among Federal Agencies
Could Reduce Unidentified Earnings Reports' which was released on March
7, 2005.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as part
of a longer term project to improve GAO products' accessibility. Every
attempt has been made to maintain the structural and data integrity of
the original printed product. Accessibility features, such as text
descriptions of tables, consecutively numbered footnotes placed at the
end of the file, and the text of agency comment letters, are provided
but may not exactly duplicate the presentation or format of the printed
version. The portable document format (PDF) file is an exact electronic
replica of the printed version. We welcome your feedback. Please E-mail
your comments regarding the contents or accessibility features of this
document to Webmaster@gao.gov.
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. 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.
Report to Congressional Committees:
United States Government Accountability Office:
GAO:
February 2005:
Social Security:
Better Coordination among Federal Agencies Could Reduce Unidentified
Earnings Reports:
GAO-05-154:
GAO Highlights:
Highlights of GAO-05-154, a report to congressional committees:
Why GAO Did This Study:
Each year, the Social Security Administration (SSA) receives millions
of employer-submitted earnings reports (Form W-2s) that it is unable to
place in an individual Social Security record. If the Social Security
number (SSN) and name on a W-2 do not match SSA‘s records, the W-2 is
retained in the Earnings Suspense File (ESF). SSA‘s ability to match
earnings reports is essential to calculating Social Security benefits.
Because of concerns about the size of the ESF, GAO was asked to
determine (1) how SSA processes workers‘ earnings reports, (2) the
types of errors in ESF reports and the characteristics of employers
whose reports are in the ESF, (3) how often earnings from repeatedly
used SSNs have been reinstated and who receives the earnings from
theses reports, and (4) what key factors contribute to ESF postings.
What GAO Found:
Upon receiving over 250 million earnings reports annually from
employers, SSA uses various processes to post such reports to workers‘
Social Security records. For reports in which worker names and SSNs
exactly match SSA‘s information, the earnings are credited to the
appropriate Social Security record. When SSA encounters earnings
reports that do not match its records, SSA attempts to make a match
through various automated processes. Such processes have allowed SSA to
identify valid records for an average of 15 million reports annually.
However, about 4 percent of the reports still remain unmatched and are
retained in the ESF. SSA uses additional automated and manual processes
to continue to identify valid records. The most recent data show that
SSA posted (’reinstated“) over 2 million earnings reports in the ESF to
valid records from such processes.
Earnings reports in the ESF have serious data problems and are
particularly likely to be submitted by certain categories of employers.
Such problems include missing SSNs and employer use of the same SSN for
more than one worker in the same tax year. Additional problems include
missing surnames or names that include nonalphabetic characters. Forty-
three percent of employers associated with earnings reports in the ESF
are from only 5 of the 83 broad industry categories. Among these
industry categories, a small portion of employers account for a
disproportionate number of ESF reports.
SSA has reinstated a substantial number of earnings reports with SSNs
that appear repeatedly in the ESF. We analyzed the most frequently
occurring 295 SSNs, which appeared in ESF 1,000 times or more between
tax years 1985 and 2000. Of the earnings reports associated with these
SSNs, SSA reinstated 13.1 million to the records of about 11.7 million
workers. Although most reinstatements were for U.S.-born workers, in
recent years the percentage of reinstatements to foreign-born workers
has markedly increased. Also increasing is the percentage of foreign-
born workers that received reinstatements for earnings in years prior
to receiving a valid SSN”a potential indicator of unauthorized
employment.
Three major factors contribute to ESF postings. Under IRS regulations,
employers must ask new hires to provide their name and SSN, but are not
required to independently corroborate this information with SSA. DHS
requires employers to visually inspect new workers‘ identity and work
authorization documents, but employers do not have to verify these
documents, and they can be easily counterfeited. Further, IRS
regulations are minimal; IRS has no record of assessing a penalty for
filing inaccurate earnings reports; and DHS enforcement efforts against
employers who knowingly hire unauthorized workers has been limited in
recent years because of shifting priorities following the events of
September 11, 2001. Last, although SSA and DHS offer employers
verification free of charge, these services are voluntary, have some
limitations, and remain underutilized.
What GAO Recommends:
GAO recommends that the Commissioners of SSA and the Internal Revenue
Service (IRS) and the Secretary of the Department of Homeland Security
(DHS) work on several fronts to facilitate more accurate earnings
reporting by employers, enhance verification systems, and institute
effective data sharing to deter unauthorized work activity and ESF
postings. SSA, IRS, and DHS agreed to consider our recommendations,
provided clarifying information, and described initiatives planned or
under way to enhance earnings reporting and worker verification.
www.gao.gov/cgi-bin/getrpt?GAO-05-154.
To view the full product, including the scope and methodology, click on
the link above. For more information, contact Barbara D. Bovbjerg at
(202) 512-7215 or bovbjergbj@gao.gov.
[End of section]
Contents:
Letter:
Results in Brief:
Background:
SSA Uses Electronic and Manual Processes to Match Earnings Reports to
Appropriate Records:
ESF Earnings Reports Frequently Include Inaccurate and Missing
Information:
Earnings from Frequently Used SSNs Are Often Reinstated and
Increasingly Belong to Foreign-Born Workers:
Several Key Factors Contribute to ESF Postings:
Conclusions:
Recommendations:
Agency Comments and Our Evaluation:
Appendix I: Objectives, Scope, and Methodology:
Appendix II: Comments from the Social Security Administration:
Appendix III: Comments from the Internal Revenue Service:
Appendix IV: Comments from the Department of Homeland Security:
Appendix V: GAO Contacts and Staff Acknowledgments:
GAO Contacts:
Staff Acknowledgments:
Tables:
Table 1: Number of Earnings Reports and Related Amounts in the ESF, by
Decade:
Table 2: Number of Earnings Reports with Initially Invalid Identity
Information That SSA Corrected Via the Upfront Validation Routines
(1998-2002):
Table 3: Number of Earnings Reports That SSA Reinstated to Correct
Records Using Selected Back-End Routines for Tax Year 2001:
Table 4: Use of One SSN by an Employer for Multiple Workers on Earnings
Reports in a Tax Year (1985-2000):
Table 5: Employers with Earnings Reports in the ESF Using One SSN More
than Once in a Tax Year (1985-2000):
Table 6: Range of Number of Employers' Earnings Reports Recorded in the
ESF, as of January 2003 (Tax Years 1985-2000):
Table 7: Ranges of Unposted Earnings Associated with the 9.58 Million
Items in the ESF under the 295 SSNs Analyzed:
Table 8: Percentage of Reinstatement Recipients Reviewed Who Were
Foreign-Born, by Year:
Table 9: Analysis of Foreign-Born Persons Receiving Reinstatements from
Repeatedly Used SSNs, Grouped by Countries of Birth:
Table 10: Percentage of Reinstatements to Foreign-Born Persons with
Earnings before Receipt of SSN, by Year of Earnings:
Figures:
Figure 1: Flowchart of SSA's Process for Posting Earnings Reports to
Workers' Social Security Records:
Figure 2: Employers Reporting Earnings in the ESF, by Type of Business:
Figure 3: Employer Investigations and Notices of Intent to Fine Letters
Issued, Fiscal Years 1992 to 2002:
Figure 4: Investigative Work Years and Arrests of Unauthorized Workers,
Fiscal Years 1999 to 2003:
Abbreviations:
DHS: Department of Homeland Security:
ESF: Earnings Suspense File:
ESLO: Employer Service Liaison Officer:
ICOR: Item Correction:
IIRIRA: Illegal Immigration Reform and Immigrant Responsibility Act:
IRCA: Immigration Reform and Control Act:
IRS: Internal Revenue Service:
MEF: Master Earnings File:
OIG: Office of the Inspector General:
PEBES: Personal Earnings and Benefit Estate Statements:
SSA: Social Security Administration:
SSN: Social Security number:
United States Government Accountability Office:
Washington, DC 20548:
February 4, 2005:
The Honorable F. James Sensenbrenner, Jr.:
Chairman:
Committee on the Judiciary:
House of Representatives:
The Honorable Jim McCrery:
Chairman:
Subcommittee on Social Security:
Committee on Ways and Means:
House of Representatives:
The Honorable E. Clay Shaw, Jr.:
Committee on Ways and Means:
House of Representatives:
The Social Security Administration (SSA) manages benefit programs that
cover nearly all working Americans. To carry out this responsibility,
SSA must maintain records of workers' earnings because they are the
basis for calculating benefits for workers and their dependents. Each
year, SSA receives over 250 million earnings reports (Form W-2s). SSA
posts earnings to its records based on the name and Social Security
number (SSN) submitted with each earnings report. If the SSN or name
does not match SSA's records, the reported earnings amount is placed in
an Earnings Suspense File (ESF), where it remains until SSA obtains
evidence to link the unidentified earnings to a valid SSN--a process
termed "earnings reinstatement." Earnings reports with incorrect or
incomplete information have been a long standing problem, and in 2002,
SSA was unable to post to valid worker records almost 9 million reports
representing $56 billion in earnings. In November 2004, SSA reported
that the ESF contained a combined total of 246 million earnings records
from all tax years back to the inception of the Social Security program
(1937), representing about $463 billion in reported earnings. Because
of your concerns about the size of the ESF, and the potentially
negative consequences for benefit payments and tax administration
associated with incorrect earnings records, you asked us to determine
(1) how SSA processes workers' earnings reports, (2) the types of
errors in ESF reports and the characteristics of employers whose
reports are in the ESF, (3) how often earnings from repeatedly used
SSNs have been reinstated and who receives the earnings from these
reports, and (4) what key factors contribute to ESF postings.
To complete our work, we met with SSA officials to obtain information
on the various electronic processes SSA uses to resolve errors and post
earnings to worker records. We also visited a total of eight SSA field
offices located in New York, New Jersey, Virginia, and California that
had significant reinstatement activity to observe and document SSA's
manual procedures for reviewing and validating evidence submitted by
individuals seeking reinstated earnings. To determine the
characteristics of ESF earnings and assess reinstatement activity, we
obtained and analyzed an electronic copy of the ESF for tax years 1985
to 2000 (84.6 million records). As part of our analysis, we assessed
the reliability of the ESF by reviewing recent SSA, Inspector General,
and contractor reports; interviewing agency officials; and checking the
databases for consistency. To analyze reinstatement of earnings
reported under repeatedly used SSNs, we selected 295 SSNs that appeared
most frequently in the ESF--each appeared 1,000 or more times. We
tracked the number of reinstatements associated with these SSNs since
the inception of the program and examined the characteristics of the
workers whose earnings had been reinstated. We also documented the
tools SSA makes available to employers to assist them in verifying
SSNs, reviewed Internal Revenue Service (IRS) regulations requiring
employers to solicit and document worker names and SSNs, and
interviewed key officials responsible for enforcing those regulations.
We also reviewed reports on the ESF prepared by SSA's Office of the
Inspector General (OIG) and private contractors. Finally, we met with
officials from the Department of Homeland Security (DHS) to document
the agency's efforts to enforce laws that prohibit employers from
knowingly hiring workers who are not authorized to work in the United
States and to obtain information about DHS's worker verification
service. Our work was conducted between October 2002 and December 2004
in accordance with generally accepted government auditing standards.
(See app. I for more details about our methodology and scope.)
Results in Brief:
SSA uses various processes to post earnings reports to workers' Social
Security records. For reports where worker names and SSNs exactly match
SSA's information, the earnings reports are credited to the appropriate
Social Security record. About 10 percent of initial submissions,
however, do not match SSA' s records, and the agency attempts to obtain
a match through automated processes that it refers to as "validation
routines." For each of these earnings reports, the agency conducts over
20 validation routines that address possible errors in worker names,
such as spelling or in SSN numbering sequence. Such routines have
allowed SSA to identify the correct SSN record and post an average of
15 million reports annually over the past 5 years. However, almost 9
million reports (about 4 percent of all reports) remain unmatched even
after the above validation routines and are, therefore, placed in the
ESF. Subsequently, SSA employs additional automated and manual
routines, including requesting updated information from workers and
employers to reconcile the earnings reports. Some of these processes
take place in SSA field offices. For example, SSA field staff may
interview workers seeking reinstatements, review the evidentiary
documents submitted, and credit the reported earnings to their Social
Security records. The most recent reinstatement data available show
that for tax year 2001 earnings reports, SSA posted about 600,000
earnings reports from the ESF, representing about $4 billion in
reported earnings. Other processes applied to a range of filing years
have recently reinstated an additional 1.9 million reports.
Many of the earnings reports in the ESF that we examined have serious
data problems and are particularly likely to be submitted by certain
types of employers. Of the 84.6 million records placed in the ESF for
tax years 1985-2000, about 9 million had all zeros (e.g., 000-00-0000)
in the SSN field. For 3.5 million records, employers used the same SSN
to report earnings for multiple workers in a single tax year. About 1.4
million records had SSNs that had never been issued by SSA, and over
260,000 were missing a first name. Certain types of employers were most
frequently associated with incorrect earnings reports. For example,
limited data provided by SSA show that 43 percent of the employers with
earnings reports in the ESF are from only 5 of the 83 broad industry
categories, with "eating and drinking establishments" and "construction
and special trades" being the top 2. A small portion of employers also
account for a disproportionate number of ESF reports. For example, only
about 8,900 employers--0.2 percent of all employers with reports
recorded in the ESF for tax years 1985-2000--submitted over 30 percent
of the reports we analyzed.
SSA eventually reinstated a substantial number of earnings reports
associated with SSNs that appeared repeatedly in the ESF; a growing
number of reinstatements have been to the Social Security records of
foreign-born workers. We analyzed 295 distinct SSNs that appeared in
ESF 1,000 times or more between tax years 1985 and 2000. Of the
earnings reports associated with these SSNs, SSA reinstated 13.1
million to the records of about 11.7 million individuals. Historically,
most workers receiving reinstatements were U.S.-born males. However, in
more recent years, the percentage of foreign-born workers receiving
such reinstatements has markedly increased, from about 8 percent before
1986 to almost 21 percent in 2003. Workers born in Mexico, Canada,
Germany, and Cuba represented nearly 40 percent of all foreign-born
individuals receiving such reinstatements. Further, in 2003 about 47
percent of foreign-born workers with reinstatements from repeatedly
used SSNs had earnings prior to actually receiving a valid SSN--a
potential indicator of unauthorized employment.
Limited requirements for obtaining and reporting accurate worker names
and SSNs, IRS enforcement weaknesses and limited DHS worksite
enforcement efforts, and underutilization of employee verification
systems create an environment in which false names and SSNs can be used
for employment purposes, and in these cases earnings records frequently
cannot be associated with the correct Social Security record. IRS
requires employers to solicit newly hired workers' names and SSN
information on IRS Form W-4 (Employee's Withholding Allowance
Certificate) that workers must complete. However, employers rely
primarily on workers to self-report this information, with no
independent corroboration by employers. While DHS's employer reporting
requirements are somewhat more demanding than IRS's, employers still
rely primarily on visual inspections of documents demonstrating
identity and work authorization, such as driver's licenses, Social
Security cards, and immigration documents to establish identity and
work authorization. Such documents can be easily counterfeited, and DHS
lacks reasonable assurance that employers will detect individuals using
them. Further, IRS's reporting requirements are minimal, and IRS has no
record of ever assessing a penalty against an employer for filing
inaccurate earnings reports. DHS enforcement activities and sanctions
against employers who hire unauthorized workers are also limited.
Following the events of September 11, 2001, DHS has focused its limited
worksite enforcement efforts primarily on critical infrastructure
facilities, such as airports and power plants, where unauthorized
workers could pose security threats, rather than on industries most
commonly associated with ESF postings. Further, although earnings
reports may involve illegal work activity, DHS has not used data
regularly supplied by SSA on potential unauthorized workers and their
employers. Although both SSA and DHS offer employers free verification
services, these services are voluntary, have some limitations, and
remain underutilized.
We are recommending that the Commissioners of the Social Security
Administration and the Internal Revenue Service and the Secretary of
Homeland Security work collaboratively on several fronts to facilitate
more accurate earnings reporting by employers, enhance existing
electronic verification systems, and institute more effective data
sharing to deter erroneous earnings postings and unauthorized SSN use.
SSA, IRS, and DHS generally agreed with our recommendations, provided
additional clarifying information regarding their roles in the current
process for soliciting and verifying worker information, and noted
various initiatives that are either planned or under way to address our
recommendations. Their comments are reproduced in appendixes II, III,
and IV.
Background:
Social Security benefits are based on a worker's lifetime earnings in
covered employment. As the agency responsible for issuing SSNs and
paying retirement, survivors, and disability benefits to insured
persons, SSA must have accurate records of every worker's earnings.
Inaccurate earnings records can create benefit payment errors.
Through a process known as enumeration, SSA assigns a unique SSN to
each individual who meets the requirements for one. Currently, SSNs are
issued to most U.S. citizens at birth. They are also available to
noncitizens lawfully admitted to the United States with permission to
work. Lawfully admitted noncitizens may also qualify for an SSN for
nonwork purposes when a federal, state, or local law requires that they
have an SSN to obtain a particular public benefit or service. SSA must
obtain documentary evidence from such applicants regarding their age,
identity, U.S. citizenship, or lawful alien status, and if they were
previously assigned an SSN. Thus, SSA maintains a historical record of
each worker's annual earnings, which is identified by the worker's name
and Social Security number.[Footnote 1]
The earnings reporting process begins at the end of each calendar year,
when employers submit reports of their workers' earnings to SSA on IRS
Form W-2 (Wage and Tax Statement). To prepare the W-2, employers
generally use certain information that workers provide on Form W-4,
which is the document that determines the amount of federal income
taxes that will be withheld from the worker's pay. If the SSN and name
on an earnings report submitted by the employer do not match
information in SSA's Master Earnings File (MEF), the reported earnings
are placed in the ESF, which is a repository for earnings reports for
unidentified workers. The ESF is an online file that can be updated
throughout the day by all SSA field offices and various centralized
components, although the updates are performed via batch mode. Removal
of earnings reports from the ESF occurs only when a report can be
matched and posted to a worker's MEF record. This process is termed
"reinstatement" by SSA. Thus, the number of reports in ESF on a given
day fluctuates as earnings are reinstated to the correct Social
Security records. Table 1 reflects the ESF reports remaining, listed by
decade, corresponding to the tax year of earnings for which each report
applied, since inception of the Social Security program.
Table 1: Number of Earnings Reports and Related Amounts in the ESF, by
Decade:
Dollars in billions.
Decade: 1937-39[A];
Number of reports: 8,908,235;
Uncredited earnings: $0.6.
Decade: 1940-49;
Number of reports: 19,764,525;
Uncredited earnings: $2.0.
Decade: 1950-59;
Number of reports: 22,155,420;
Uncredited earnings: $3.6.
Decade: 1960-69;
Number of reports: 28,294,126;
Uncredited earnings: $6.7.
Decade: 1970-79;
Number of reports: 44,402,863;
Uncredited earnings: $22.7.
Decade: 1980-89;
Number of reports: 41,928,484;
Uncredited earnings: $77.3.
Decade: 1990-99;
Number of reports: 51,950,009;
Uncredited earnings: $188.9.
Decade: 2000-02[A];
Number of reports: 28,339,912;
Uncredited earnings: $161.0.
Total;
Number of reports: 245,743,574;
Uncredited earnings: $462.8.
Source: GAO analysis of SSA data.
[A] In November 2004, SSA provided us with a summary of the items in
the ESF showing the number of earnings reports and their dollar amount
remaining in the ESF by year of earnings. The data reflect only partial
decade information for indicated years.
[End of table]
SSA Uses Electronic and Manual Processes to Match Earnings Reports to
Appropriate Records:
SSA uses various processes to post reported workers' earnings to valid
Social Security records. Generally, employers send SSA one W-2 each
year that reports the annual earnings for each of their workers. Upon
receipt of these earnings reports, SSA electronically validates whether
it has established a Social Security record for the reported SSN and
surname shown on the W-2. SSA does this by electronically matching the
worker's surname and SSN on the W-2 to information in its number
identification file (Numident) that contains demographic information
about every SSN holder. When the SSN and the first seven characters of
the surname are identical on the W-2 and the Numident file, SSA posts
those earnings to the indicated record in its MEF. SSA is able to place
about 90 percent of employer-submitted earnings reports received each
year in an appropriate MEF record. (Fig. 1 shows SSA's process in more
detail.)
Figure 1: Flowchart of SSA's Process for Posting Earnings Reports to
Workers' Social Security Records:
[See PDF for image]
[End of figure]
For the 10 percent of the reports that fail this initial validation
test, SSA performs more than 20 of what it calls "front-end validation
routines" that manipulate either the reported name or the SSN in a
variety of ways to correct common reporting mistakes so that SSA can
find an MEF record and prevent the posting of the earnings to the ESF.
SSA's front-end routines identify Social Security records for about 60
percent of the reports that are initially categorized as mismatches
each year. In manipulating worker information to find a valid record,
the automated front-end routines assume either the SSN is correct and
there is a problem in the reported name or vice versa. For example, one
front-end routine for reported name errors tests whether the first name
and the surname have been reversed on the employer-filed W-2. The name
reversal routine compares the first name from SSA's Numident file with
the surname on the W-2. If they are the same and the first initial for
the middle and surname match the information on the W-2, then SSA
assumes it has found the proper record and posts the earnings. Other
front-end routines check whether digits in the SSN are transposed or
inaccurate or whether the name on the report contains transposed or
missing letters. Another front-end routine involves searching for
reinstated items that in the past have included the same reporting
error occurring in the current year's earnings report.
Over the past several years, such routines have allowed SSA to post an
annual average of 15 million earnings reports to individual MEF
records, rather than to the ESF. Table 2 summarizes the performance of
these front-end routines for the past 5 reporting years. It shows that
SSA found about 76 million valid records for reported earnings for tax
years 1998 to 2002.
Table 2: Number of Earnings Reports with Initially Invalid Identity
Information That SSA Corrected Via the Upfront Validation Routines
(1998-2002):
Year earnings were processed: 1998;
Millions of reports not initially matching SSA records: 23.9.
Year earnings were processed: 1999;
Millions of reports not initially matching SSA records: 25.8.
Year earnings were processed: 2000;
Millions of reports not initially matching SSA records: 28.4.
Year earnings were processed: 2001;
Millions of reports not initially matching SSA records: 24.4.
Year earnings were processed: 2002[A];
Millions of reports not initially matching SSA records: 21.9.
Total;
Millions of reports not initially matching SSA records: 124.4.
Year earnings were processed: 1998;
Millions of corrected reports: 15.9.
Year earnings were processed: 1999;
Millions of corrected reports: 16.7.
Year earnings were processed: 2000;
Millions of corrected reports: 17.6.
Year earnings were processed: 2001;
Millions of corrected reports: 13.6.
Year earnings were processed: 2002[A];
Millions of corrected reports: 12.0.
Total;
Millions of corrected reports: 75.8.
Percentage corrected;
Year earnings were processed: 1998;
Percentage corrected: 66.4%.
Year earnings were processed: 1999;
Percentage corrected: 64.7%.
Year earnings were processed: 2000;
Percentage corrected: 62.1%.
Year earnings were processed: 2001;
Percentage corrected: 56.0%.
Year earnings were processed: 2002[A];
Percentage corrected: 54.8%.
Total;
Percentage corrected: 61.0%.
Source: GAO analysis of SSA data.
[A] Data as of November 2004.
[End of table]
If the front-end routines do not identify a valid record, SSA posts the
earnings in the ESF. SSA subsequently performs what it calls "back-end"
processes on the items, consisting of electronic and manual actions to
match the earnings to a worker's MEF record. For one such process, SSA
uses corrections of reported names and SSNs generated under IRS's
automated routines. SSA then attempts to find the W-2 in the ESF and
validate the corrected name and SSN that IRS provides against SSA's
records; when both conditions are met, SSA accepts IRS's corrections
and reinstates the item to the worker's record. Under another process,
which SSA calls decentralized correspondence, or DECOR, SSA sends
letters to addresses listed on each invalid W-2, seeking information to
resolve the identity issue (or to the employer if the W-2 lacks a valid
address). If the worker does not respond, SSA then sends a letter to
the employer that filed the report soliciting assistance in resolving
the problem. Other types of correspondence involve Young Child Earnings
Records and Earnings After Death, for which SSA sends letters to
employers and/or the persons whose SSNs appear on the reports; SSA
automatically posts such earnings reports to the ESF because the
persons named in the reports are, respectively, either (a) 6 years of
age or younger (thus unlikely to have earnings through employment) or
(b) have a date of death recorded on their Numident record for a year
prior to the tax year for which earnings on Form W-2 have been
reported. Upon receiving satisfactory documentation clarifying the
earnings and linking them to the proper SSN, SSA reinstates earnings
reports to the individuals' MEF record.
SSA also uses yet another process, known as FERRET, that compares
worker addresses on the W-2 with addresses that IRS has from individual
tax returns. In its Operation 30 routine, SSA staff compare ESF
earnings reports with valid SSNs with information in the Numident
record. Staff check whether nicknames, surname derivations, and other
obvious mistakes in spelling might be the cause of the posting problem.
Table 3 shows that in 2001 (the last year for which data were
available), selected back-end routines reinstated almost 600,000
earnings reports totaling almost $4 billion.
Table 3: Number of Earnings Reports That SSA Reinstated to Correct
Records Using Selected Back-End Routines for Tax Year 2001:
Dollars in millions.
Back-end reinstatement routine: IRS corrections (mismatch resolutions
by IRS sent to SSA);
Items reinstated: 398,307;
Earnings reinstated: $2,608.7.
Back-end reinstatement routine: FERRET;
Items reinstated: 97,145;
Earnings reinstated: $557.3.
Back-end reinstatement routine: W-2 corrections submitted by employers;
Items reinstated: 60,791;
Earnings reinstated: $529.7.
Back-end reinstatement routine: Operation 30;
Items reinstated: 29,661;
Earnings reinstated: $202.2.
Back-end reinstatement routine: Correspondence routines: DECOR;
Items reinstated: 2,980;
Earnings reinstated: $20.6.
Back-end reinstatement routine: Correspondence routines: Young child
earnings records;
Items reinstated: 499;
Earnings reinstated: $5.1.
Back-end reinstatement routine: Correspondence routines: Earnings after
death;
Items reinstated: 127;
Earnings reinstated: $1.3.
Total;
Items reinstated: 589,510;
Earnings reinstated: $3,924.8.
Source: SSA.
[End of table]
SSA has two other electronic back-end routines that have produced a
large number of reinstatements. In a process called SWEEP, SSA
periodically reruns ESF items through its records to determine whether
updated information has been added to the Numident or whether newly
developed validation routines might permit reinstatements. In 2003, SSA
reinstated 123,741 items through SWEEP, covering tax years 1977-2001.
GAP SWEEP is a newly developed routine that scans earnings records for
valid SSNs in the ESF and assesses whether yearly gaps in earnings
exist in the MEF record and might be linked to similar earnings in the
ESF.[Footnote 2] If a link can be made, SSA uses slightly less
stringent name match rules; if the name can be validated, the item is
reinstated. As of May 2003, SSA has reinstated through the GAP SWEEP
routine over 1.5 million items (across all tax years back to 1937),
representing $6.1 billion in earnings.
Still another back-end process involves a manual review of worker-
submitted evidence and a check of automated data. Workers (and their
dependents and survivors) may visit local SSA offices to have earnings
reinstated through the Item Correction (ICOR) process. Individuals
provide SSA staff with evidence, such as W-2s, earnings statements, and
tax returns, to document earnings that are missing from their Social
Security record. Upon receiving adequate proof that links an earnings
report to the individual, SSA field staff manually reinstate the
earnings, subject to an accuracy check by a peer or supervisor. SSA
provided information indicating that in fiscal year 2003, field staff
had made about 244,000 earnings reinstatements through the ICOR process.
Furthermore, each year SSA mails a Social Security statement to workers
and former workers age 25 and over who are not yet receiving
benefits.[Footnote 3] The statement lists the amount of earnings posted
to the person's Social Security record by year and encourages persons
to contact SSA about any missing earnings. Such earnings might have
been placed in the ESF because of a name or SSN mismatch.
Reinstatements related to Social Security statements are included in
the ICOR data discussed above.
ESF Earnings Reports Frequently Include Inaccurate and Missing
Information:
Earnings reports in the ESF have serious data problems. Such data
problems include missing SSNs or names, never issued numbers, and
employer use of the same SSN to report earnings for multiple workers in
a single tax year. In addition, a small portion of employers account
for a disproportionate number of ESF reports, and employers in certain
industry categories are more likely than others to submit reports with
invalid worker identity information.
ESF Reports Have Serious Data Problems:
Out of the 84.6 million reports in the ESF for the 16 tax years that we
examined (1985-2000), some of the more serious or obvious problems were
that:
* 8.9 million had all zeros in the SSN field[Footnote 4] and:
* 1.4 million had reported SSNs that were never issued.
In addition, over 270,000 of the reports had various name problems. For
example,
* 60,476 had no surname;
* 261,744 had no first name;
* 3,760 reports contained nonalphabetic characters in the name field,
such as ?, /, %,
