CONGRESSIONAL RESPONSE REPORT

Mission

We improve SSA programs and operations and protect them against fraud, waste, and abuse by conducting independent and objective audits, evaluations, and investigations. We provide timely, useful, and reliable information and advice to Administration officials, the Congress, and the public.

Authority

The Inspector General Act created independent audit and investigative units, called the Office of Inspector General (OIG). The mission of the OIG, as spelled out in the Act, is to:

Conduct and supervise independent and objective audits and investigations relating to agency programs and operations.
Promote economy, effectiveness, and efficiency within the agency.
Prevent and detect fraud, waste, and abuse in agency programs and operations.
Review and make recommendations regarding existing and proposed legislation and regulations relating to agency programs and operations.
Keep the agency head and the Congress fully and currently informed of problems in agency programs and operations.

To ensure objectivity, the IG Act empowers the IG with:

Independence to determine what reviews to perform.
Access to all information necessary for the reviews.
Authority to publish findings and recommendations based on the reviews.

Vision

By conducting independent and objective audits, investigations, and evaluations, we are agents of positive change striving for continuous improvement in the Social Security Administration's programs, operations, and management and in our own office.

December 18, 2006

The Honorable Jim McCrery
Chairman, Subcommittee on Social Security
Committee on Ways and Means
House of Representatives
Washington, D.C.  20515

Dear Mr. McCrery:

I am pleased to provide you with the enclosed report addressing your April 7, 2006 letter, requesting an assessment of the accuracy of Social Security Administration’s (SSA) Numident fields that are relied on by the Basic Pilot—a Department of Homeland Security (DHS) employment eligibility verification service. 

Thank you for bringing your concerns to my attention.  My office is committed to combating fraud, waste, and abuse in SSA’s operations and programs.  The report highlights various facts pertaining to the issues raised in your letter.  To ensure SSA, DHS and the Department of State are aware of the information provided to your office, we are forwarding a copy of this report to the Inspectors General for each of these agencies. 

If you have any questions concerning this matter, please call me or have your staff contact H. Douglas Cunningham, Assistant Inspector General for Congressional and Intra‑Governmental Liaison, at (202) 358-6319.

Sincerely,

Patrick P. O’Carroll, Jr.
Inspector General

Enclosure

cc: 

Jo Anne B. Barnhart

Executive Summary

OBJECTIVE
To assess the accuracy of the Social Security Administration’s (SSA) Numident fields that are relied on by the Basic Pilot.

background

As of November 2005, SSA had assigned about 435 million Social Security numbers (SSN)—for the primary purpose of accurately reporting and recording the earnings of people who work in jobs covered by Social Security.  When SSA assigns an SSN to an individual, the Agency creates a master record in its “Numident” file containing relevant information about the numberholder. 

SSA provides employers with a number of verification programs and services that allow them to match employees’ names and SSNs with SSA’s records.  Additionally, the Basic Pilot, a Department of Homeland Security (DHS) program supported by SSA, provides employers a tool for determining whether newly-hired employees reported the correct name, SSN, date of birth and are authorized to work in the United States.  Recently, the House and Senate each passed immigration reform bills that would mandate employers’ use of an employment eligibility verification system similar to the Basic Pilot.  Given the extent and significance of this proposed legislation, it is essential that the Numident file contain correct information.

On April 7, 2006, we received a request from the Honorable Jim McCrery of the House Committee on Ways and Means, Chairman of the Subcommittee on Social Security, to assist the Subcommittee in obtaining information on the reliability of SSA data used in the Basic Pilot program to verify employment eligibility.  In particular, the Chairman asked that we assess the accuracy of SSA Numident fields that are relied on by the Basic Pilot.  The Chairman specified that we determine Numident accuracy rates for each of the following populations:  (1) native‑born U.S. citizens, (2) foreign‑born U.S. citizens, and (3) non‑U.S. citizens.  To address his request, we reviewed 810 randomly-selected Numident records in each of these populations (a total of 2,430 records) to determine their accuracy and whether data discrepancies might result in inaccurate feedback from the Basic Pilot.  Because SSA does not delete, destroy, rescind, inactivate or cancel SSNs once they are assigned, we selected our sample of all SSNs the Agency had assigned as of November 30, 2005.

RESULTS OF REVIEW

We cannot predict the types of documentation each of the tested numberholders might present to an employer when attempting to prove their identities and authorization to work in the United States—as the current process allows a number of varying sources of this information.   Accordingly, our audit conclusions only pertain to the accuracy of SSA’s Numident file when compared to (1) information SSN numberholders provided to SSA when applying for their original and/or replacement Social Security cards and, if applicable, (2) certain data elements DHS had for the Numident records tested.  Despite these limitations, we found SSA’s Numident file information to be generally accurate. 

We identified some discrepancies that could result in the Basic Pilot providing incorrect feedback to employers attempting to determine the employment eligibility of their workers.  Specifically, of the 2,430 Numident records reviewed, 136 contained discrepancies in the name, date of birth or citizenship status of the numberholder or we determined that the numberholder may be deceased.   In all of these cases, the Basic Pilot provided incorrect results.  As a result, we estimate that discrepancies in approximately 17.8 million (4.1 percent) of the 435 million Numident records could result in incorrect feedback when submitted through the Basic Pilot.  While the accuracy of SSA’s Numident records is noteworthy, if use of an employment verification service such as the Basic Pilot becomes mandatory, the workload of SSA and DHS may significantly increase—even if only a portion of these 17.8 million numberholders need to correct their records with one of these agencies.

misclassifications, we estimate that of the 46.5 million non-U.S. citizen records in SSA’s Numident file, about 3.3 million contain out-of-date citizenship status codes.   As such, these individuals may need to visit an SSA office to correct their Numident record before they would be confirmed eligible for employment by the Basic Pilot.

CONCLUSION

Given the scope and breadth of information held in SSA’s Numident file, we applaud the Agency on the accuracy of the data we tested.  However, we estimate that approximately 17.8 million Numident records contain discrepancies that may result in incorrect Basic Pilot feedback to employers.  As Congress considers legislation requiring mandatory verification of all U.S. workers’ employment eligibility through a system such as the Basic Pilot, we believe it should examine the significant workload that may result from the millions of numberholders whose Numident records may need to be corrected.  Our review showed that the Numident records for these individuals have discrepancies in the numberholders’ names, dates of birth, citizenship status and/or death indications that would result in inaccurate feedback from the Basic Pilot. 

Because our tests included SSNs that were assigned decades ago, we recognize that some numberholders would no longer be working and would not attempt to correct their SSA and/or immigration records.  However, if even a portion of the estimated numberholders whose Numident records contained discrepancies were required to visit an SSA office to correct their information, the Agency’s workload may significantly increase until such time as the affected records were corrected. 

Table of Contents

INTRODUCTION................................................................................................................ 1

RESULTS OF REVIEW..................................................................................................... 5

Native-Born U.S. Citizens................................................................................................... 6

Foreign-Born U.S. Citizens................................................................................................. 9

Non-U.S. Citizens.............................................................................................................. 11

CONCLUSION.................................................................................................................. 14

OTHER MATTER............................................................................................................. 15

APPENDICES

APPENDIX A – Acronyms

APPENDIX B – Scope and Methodology

APPENDIX C – Sample Results

APPENDIX D – Basic Pilot

APPENDIX E – Additional Discrepancies that Did Not Impact Basic Pilot Results

INTRODUCTION

OBJECTIVE

To assess the accuracy of the Social Security Administration’s (SSA) Numident fields that are relied on by the Basic Pilot.

BACKGROUND

As of November 2005, SSA had assigned about 435 million Social Security numbers (SSN)—for the primary purpose of accurately reporting and recording the earnings of people who work in jobs covered by Social Security.  When SSA assigns an SSN to an individual, the Agency creates a master record in its “Numident” file containing relevant information about the numberholder.  This information includes the numberholder’s name, date of birth, place of birth, parents’ names, citizenship status, date of death (if applicable) and the office where the SSN application was processed and approved.  Additionally, the Numident record for each numberholder identifies (1) any changes to the original information provided by the numberholder (for example, name changes, revisions to citizenship status) and (2) an account of all replacement SSN cards obtained. 

The first explicit statutory authority to issue SSNs was enacted in 1972.   Before that time, SSNs were issued pursuant to administrative procedures that the Agency had established.  At the inception of the program, all SSNs were assigned and cards issued based solely on information provided by the applicant.  Evidence of identity was not required.  However, beginning in November 1971, persons age 55 and over applying for an SSN for the first time were required to submit evidence of identity.  As of April 1974, noncitizens were required to submit documentary evidence of age, identity and immigration status.  In May 1978, SSA began requiring all SSN applicants to provide evidence of age, identity and U.S. citizenship or noncitizen status.  In addition, as of September 2002, SSA began verifying all immigration documents with DHS before assigning SSNs to noncitizens.

SSA posts an employee’s wages to its Master Earnings File (MEF) based on a Wage and Tax Statement (Form W-2) submitted by his or her employer.   However, these earnings can only be posted successfully to a numberholder’s MEF record if the name and SSN reported on the Form W-2 match the information on SSA’s Numident file for that individual.  For example, if the Numident record for a numberholder contains a last name that significantly differs from what the employer provided on the Form W-2, the earnings will not be posted to the individual’s earnings record—and, if not corrected, will not be counted towards the numberholder’s future Social Security benefits.  Additionally, if an employee falsely uses a name and/or SSN to work in the United States, which does not match the associated Numident record, these wages will not be posted to the correct numberholder’s account. 

To prevent these name and SSN discrepancies, SSA provides employers and third parties with a number of verification programs and services that allow them to verify employees’ names and SSNs against SSA’s records before submitting the Form W‑2s.   To ensure correct responses to employers through these verification services, it is essential that SSA’s Numident records are as accurate and complete as possible. 

On April 7, 2006, we received a request from the Honorable Jim McCrery of the House Committee on Ways and Means, Chairman of the Subcommittee on Social Security, to assist the Subcommittee in obtaining information on the reliability of SSA data the Department of Homeland Security (DHS) uses to verify employment eligibility.  In particular, the Chairman asked that we assess the accuracy of SSA Numident fields that are relied on by the Basic Pilot.  The Chairman specified that we determine Numident accuracy rates for each of the following populations:  (1) native‑born U.S. citizens, (2) foreign‑born U.S. citizens, and (3) non‑U.S. citizens.

BASIC PILOT

If SSA verifies this information and no date of death is recorded on the Numident record, SSA provides DHS with U.S. citizenship information for the individual—as annotated on the numberholder’s most recent Numident record.  If the employee alleges U.S. citizenship and the Numident record also indicates the individual is a U.S. citizen, the Basic Pilot then notifies the employer that work authorization is confirmed.  If a discrepancy is noted, the Basic Pilot sends the employer an “SSA tentative nonconfirmation” response. 

After SSA verifies the SSN, name and date of birth of noncitizens, the Basic Pilot confirms current authorization to work in the United States with DHS.   If a discrepancy occurs with an employee’s work authorization, the Basic Pilot sends the employer a “DHS tentative nonconfirmation” response.  In the case of either an SSA or DHS tentative nonconfirmation, the employee is provided an opportunity to correct the information contained in SSA and/or DHS records.  See Appendix D for more information on the Basic Pilot.

SCOPE AND METHODOLOGY

To achieve our objective, we reviewed a total of 2,430 SSNs that SSA assigned from 1936 through November 2005.  We recognize that our tests included SSNs that were assigned decades ago and some of these numberholders would no longer be working and/or attempting to correct their SSA and/or immigration records.  However, we included all SSNs the Agency had assigned up to November 30, 2005, in our universe because SSA does not delete, destroy, rescind, inactivate or cancel SSNs once they are assigned.  Accordingly, older SSNs could still be used—appropriately or inappropriately—for employment purposes. 

We selected our samples from a segment of SSA’s Numident file as of November 30, 2005.   Specifically, as shown in Table 1, we reviewed 810 Numident records for SSNs within each of the following 3 populations:  (1) native‑born U.S. citizens, (2) foreign‑born U.S. citizens, and (3) non‑U.S. citizens.

Table 1:  Details of Sample Selected for Review

Our tests compared the Numident data (name, SSN, date of birth and citizenship status) to copies of the original Application for a Social Security Card (Form SS-5) and, when applicable and available, copies of applications for replacement Social Security cards to ensure that SSA personnel accurately entered the information provided by the applicant.  Additionally, we attempted to determine the legitimacy of any dates of death present on Numident records.  Further, we attempted to verify noncitizens’ and foreign‑born U.S. citizens’ “citizenship” status with DHS or the Department of State (State).  We also submitted the 1,620 U.S. citizen SSNs through the Basic Pilot, using information provided by the numberholder on his or her Form SS‑5.  Finally, we submitted all noncitizen SSN records through the Basic Pilot when we could obtain DHS alien registration or admission numbers or when we identified any discrepancies between SSA and DHS information for the numberholder. 

Underlying Assumptions

In performing our tests, we could not predict what proof of identity and work authorization a numberholder would present to an employer.  Therefore, we made the following assumptions.

• The information provided by the numberholder on his or her Form SS-5(s) was correct and would be the same data provided to an employer. 
• The information provided by DHS and/or State was correct.

See Appendix B for more information regarding our scope and methodology and Appendix C for our sample results.

RESULTS OF REVIEW

SSA’s Numident file represents an extensive repository of personal information provided by hundreds of millions of SSN applicants throughout the Agency’s history.  The primary purpose in maintaining this information is to ensure that U.S. workers receive appropriate credit for the wages they earned and, ultimately, to provide full and accurate benefits to entitled individuals.  Recently, the House and Senate each passed immigration reform bills that would mandate employers’ use of an employment eligibility verification system similar to the Basic Pilot.   Given the extent and significance of the proposed legislation, it is essential that the Numident file contain correct information.

We cannot predict what documentation every numberholder tested might present to an employer when attempting to prove their identities and authorization to work in the United States—as the current process allows a number of varying sources of this information.  Accordingly, our audit conclusions only pertain to the accuracy of SSA’s Numident file when compared to (1) information SSN numberholders provided to SSA when applying for their original and/or replacement Social Security cards and, if applicable, (2) certain data elements DHS had for the Numident records tested.  Despite these limitations, we found SSA’s Numident file information to be generally accurate. 

We identified some discrepancies that could result in incorrect feedback to employers attempting to verify employees’ eligibility to work in the United States through the Basic Pilot.  Specifically, of the 2,430 Numident records reviewed, 136 contained discrepancies that could result in inaccurate employment eligibility results—and, in most cases, additional work for SSA and DHS as employees attempt to correct their Numident and immigration records.   As a result, we estimate that data discrepancies in approximately 17.8 million (4.1 percent) of the 435 million Numident records could result in incorrect feedback when submitted through the Basic Pilot.   These discrepancies are outlined in Table 2. 

Table 2:  Data Discrepancies in SSA’s Numident File

NATIVE-BORN U.S. CITIZENS

Generally, SSA Numident records for native-born U.S. citizens were accurate.  Of the 810 Numident records reviewed in this population, we identified 27 (3.3 percent) discrepancies that may result in inaccurate employment eligibility feedback from the Basic Pilot.  Given this accuracy rate, we estimate that of the approximately 380 million native-born U.S. citizen records in the Numident file, about 12.7 million contain discrepant information that may result in incorrect Basic Pilot feedback.  As shown in Table 3, these discrepancies involved name, date of birth, and death information that did not match or had not been annotated on SSA’s Numident file.  Additionally, we noted that 19 (2.3 percent) of the 810 U.S. born citizen records we tested were age 90 and over, had no indication of death on any SSA record and were not currently receiving SSA benefits.  While these individuals may still be alive and, in some rare cases working, the Basic Pilot has no edit that would require additional contact with these individuals before confirming employment eligibility.  We encourage SSA and DHS to incorporate a step in the Basic Pilot requiring further verification of potential employees who have reached a pre-determined age and/or have some indication of death on an SSA system of record.

Table 3:  Native-Born U.S. Citizen Record Discrepancies that May Affect Basic Pilot Results

Name and Date of Birth Discrepancies that Affected Basic Pilot Results

Two (0.24 percent) of the 810 native-born U.S. citizen Numident records tested contained inaccurate spellings of the numberholders' names—when compared to the associated Form SS-5s.  When the proper names were tested in the Basic Pilot, it returned an “SSA tentative nonconfirmation” for each.  Additionally, 2 (0.24 percent) other Numident records lacked a date of birth for the numberholders.  When we tested the dates of birth provided on the numberholders’ Form SS-5s, the Basic Pilot returned an “SSA tentative nonconfirmation” response.  As such, these individuals’ employment eligibility may not be initially verified through the Basic Pilot and the numberholders may need to correct their information with SSA. 

Death Indication Discrepancies that May Affect Basic Pilot Results

SSA provides a death indicator in its responses to the Basic Pilot.  However, we determined that 23 (2.8 percent) of the 810 native-born U.S. citizens within our sample may be deceased according to SSA records, yet the Basic Pilot confirmed employment eligibility for these numberholders.  In seven of these cases, the Numident records had no indication that the numberholder was deceased, but other SSA records (that is, the Master Beneficiary Record or Supplemental Security Income Record ) provided a date of death for the individuals.  In the other 16 cases, the Numident records were annotated with a transaction code of “D,” which indicates that a report of death was received by SSA, but the death was not confirmed.  

Although SSA may not have verified the deaths of these numberholders, we do not believe the Basic Pilot should confirm employment eligibility without some additional contact with these individuals.  If someone presented themselves as one of these 23 individuals to an employer and provided the correct name, SSN and date of birth, the employer could incorrectly verify the individual’s employment eligibility through the Basic Pilot.  Accordingly, we encourage SSA to reconcile the Numident with other SSA records and annotate Numident records when an indication of death is present.  Additionally, we believe the Basic Pilot should issue a tentative nonconfirmation notice to employers that more investigation is necessary for individuals with a transaction code of “D” on the Numident record (and any other code SSA chooses to use for death indicators annotated from the aforementioned reconciliation).  These suggestions are not intended to preclude verification entirely, but to require further inquiry and resolution before DHS confirms employment eligibility.

In response to a draft of this report, SSA requested that we note the following. 

SSA is working with State governments to improve the current paper-based process which we believe will resolve some of the issues with the Numident record not being updated.  However, the most efficient manner to improve timeliness and accuracy of State data is by using Electronic Death Reporting (EDR); a web-based automation of the death registration process.  Our goal is to receive verified death data within 24 hours of receipt in the State repository and within 5 days of death.  To achieve that goal we have currently awarded 31 State EDR development contracts.  Fourteen States plus the District of Columbia and the jurisdiction of New York City have already implemented EDR.  We expect that an additional 7 States in FY 2007 and 8 more States in FY 2008 should implement EDR.  Additionally, for the longer term, the provisions of the Intelligence Reform and Terrorism Prevention Act (IRTPA) of 2004 give grants to the States to cross match birth and death records within and among the States, and give the lead for the funding of EDR to the Department of Health and Human Services (HHS).  We continue to work with HHS for the transition of funding for EDR beginning this fiscal year.

Numberholders Age 90 and Over Receive No Further Screening by Basic Pilot

Of the 810 native-born U.S. citizens reviewed, we identified 19 (2.3 percent) individuals age 90 and over who had no indication of death on his or her Numident record or any other SSA system of record and were not currently receiving SSA benefits.  The oldest of these numberholders would be 134 in 2006.  While these are clearly old Numident records established before SSA implemented improved procedures for reporting and verifying deaths, we are concerned that the Basic Pilot has no control requiring additional investigation or contact with these numberholders before confirming employment eligibility.  In fact, we tested these 19 individuals through the Basic Pilot and all were confirmed eligible for employment. 

We acknowledge that an employer should raise their own concerns when an employee claims to have been born in 1872.  However, to ensure that (1) no employee attempts to misuse the identity of a deceased or aged person and (2) the employer does not inadvertently or intentionally overlook the misuse of the aged person’s name, SSN and date of birth, we believe the Basic Pilot should be revised to issue a notice to employers that more investigation is necessary when an employee is older than a pre‑established age.

Informal comments SSA provided in response to a draft of this report pointed out that it is illegal for employers to discriminate on the basis of age and that use of an indicator for individuals over a specified age would require a new or amended routine under the Privacy Act of 1974 to permit disclosure.  Further, the Agency stated that its Memorandum of Understanding with DHS would need to be modified to provide such an indicator.  Finally, SSA pointed out that DHS Form I-9 already requires employers to certify that the documents submitted by an employee appear to be genuine and relate to that employee. 

While we are sensitive to the possibility of age discrimination and understand the legal and contractual modifications that would be required, we continue to believe that the Basic Pilot should not routinely confirm employment eligibility for a 134-year old person—without, at the very minimum, some further conversation with that individual.  The determination whether such a control should be implemented rests with SSA, DHS and Congress.  However, we would be remiss if we did not point out this vulnerability in the current process.

FOREIGN-BORN U.S. CITIZENS

Of the 810 foreign-born U.S. citizen Numident records reviewed, 25 (3.1 percent) contained discrepant information that could result in incorrect employment eligibility feedback from the Basic Pilot.  As a result, we estimate that about 248,560 of the approximately 8 million foreign-born U.S. citizen records contain discrepancies that may result in incorrect employment eligibility results from the Basic Pilot.  As shown in Table 4, differences were noted between the name, date of birth or citizenship status SSA annotated on the 25 Numident records and data provided by (1) the numberholder on his or her Form SS-5 or (2) DHS.  Further, we also noted that SSA documented an incorrect citizenship code for 43 (5.3 percent) other foreign-born U.S. citizens within our sample (which are not listed in Table 4.)  However, these noncitizens were authorized to work in the United States, so the incorrect citizenship code did not impact the Basic Pilot results at the time of our review.

Table 4:  Foreign-Born U.S. Citizen Record Discrepancies that May Affect Basic Pilot Results

Name and Date of Birth Discrepancies that Affected Basic Pilot Results

Of the 810 foreign-born U.S. citizen Numident records tested, we identified 4 (0.5 percent) discrepancies in the “name” field and 2 (0.2 percent) differences in the “date of birth” field that caused the Basic Pilot to issue “SSA tentative nonconfirmation” responses.  In one case, when applying for an SSN, the numberholder documented on the Form SS-5 that her first name was “Sabrina.”  However, SSA recorded the numberholder’s name as “Brina” in the Numident record.  The Basic Pilot provides tolerances for some differences in the spelling of names.  However, the Basic Pilot did not recognize these two names as the same individual.  In five other cases, the name or date of birth SSA documented on the Numident record disagreed with information provided by DHS.   These discrepancies also impacted Basic Pilot results and caused “SSA tentative nonconfirmation” responses. 

Citizenship Status Discrepancies

Based on DHS information, we determined that 62 (7.7 percent) of the 810 foreign-born U.S. citizen records reviewed were not actually U.S. citizens.  Because SSA does not maintain copies of documents SSN applicants present to the Agency, we were unable to determine whether these individuals presented false evidence of U.S. citizenship to SSA or Agency personnel simply miscoded the information.   Despite the citizenship indications on SSA’s Numident records, these numberholders are not citizens and their work authorization status should be determined by DHS.  We estimate that about 616,420 other numberholders in the foreign-born U.S. citizen population (which totals approximately 8 million) may have been improperly classified as U.S. citizens in SSA records. 

Of the 62 misclassified numberholders, 43 are Legal Permanent Residents (LPR) and, as such, authorized to work in the United States.  These 43 discrepancies did not result in incorrect Basic Pilot feedback because employers would have been accurately notified that these individuals were eligible for employment.  However, if something changes with their LPR status (for example, emigration, deportation, or failure to renew their status), their permission to work in the United States may be taken away.  In fact, one DHS study estimated that, between 1900 and 1990, one-quarter to one-third of all LPRs emigrated from the United States.   Yet, the SSNs assigned to these individuals

remain active.  Therefore, the numberholders—or anyone posing as the numberholders—could continue to use the SSNs to work in the United States long after they are authorized to do so.  Accordingly, we caution SSA and DHS that improper employment eligibility confirmations may eventually result from the incorrect citizenship status currently shown on these SSA records. 

According to DHS, the remaining 19 misclassified numberholders needed further investigation before their work authorization status could be verified.  Specifically, if SSA’s citizenship codes had been correct for these numberholders, the Basic Pilot would have referred these records to DHS for review and possible contact with the numberholder before work authorization was established.  Given the uncertainty of the numberholders’ employment eligibility, we considered the discrepancies in SSA’s citizenship codes to have caused improper Basic Pilot feedback and included these cases in Table 4.

NON-U.S. CITIZENS

Non-U.S. citizen (noncitizen) records within SSA’s Numident file contained a much higher degree of error than the U.S. citizen populations.  Specifically, of the 810 noncitizen records tested, 84 (10.4 percent) contained discrepancies that may result in inaccurate employment eligibility results from the Basic Pilot.  As shown in Table 5, these discrepancies resulted from differences in names, dates of birth, citizenship status and death indications.  Many of these differences were caused because the numberholder did not update his or her information with SSA (for example, the noncitizen did not notify SSA when he or she became a U.S. citizen).  As a result, we estimate that of the approximately 46.5 million noncitizen records contained in SSA’s Numident file, 4.8 million contain discrepancies that could require the numberholder to visit an SSA office to correct his or her Numident record before employment eligibility would be confirmed. 

Additionally, as in the native‑born U.S. citizen population, we identified 50 numberholders who were age 90 or over and had no indication of death present on their SSA records and were not currently receiving SSA benefits.  As currently structured, the Basic Pilot would confirm employment eligibility for these individuals without any further contact or investigation.

Table 5:  Non-U.S. Citizen Record Discrepancies that May Affect Basic Pilot Results

Name and Date of Birth Discrepancies that Affected Basic Pilot Results

Fourteen (1.7 percent) of the 810 Numident records tested contained numberholders' names that differed from either the individuals’ Form SS-5 or DHS-provided data.   When the alternate spellings were tested in the Basic Pilot, it returned an “SSA tentative nonconfirmation” for each.  Additionally, 3 (0.4 percent) other Numident records contained a date of birth that differed from the Form SS-5 or DHS data.  As such, these individuals’ employment eligibility may not be initially verified through the Basic Pilot and the numberholders may have to correct their information with SSA. 

Citizenship Status Discrepancies that Affected Basic Pilot Results

Based on DHS information, we determined that 57 (7.0 percent) of the 810 noncitizens tested were actually U.S. citizens—but had not updated their immigration/citizenship status with SSA.  As a result, if these individuals alleged citizenship to an employer, the Basic Pilot would not initially confirm employment eligibility.  Although SSA is not at fault for these misclassifications, we estimate that of the 46.5 million non-U.S. citizen records in SSA’s Numident file, about 3.3 million contain incorrect citizenship status codes.   As such, these individuals may need to visit an SSA office to correct their Numident records before they would be confirmed eligible for employment by the Basic Pilot.
Death Indication Discrepancies that May Affect Basic Pilot Results

Additionally, we determined that 10 (1.2 percent) of the 810 noncitizens within our sample may be deceased according to SSA records.  However, the Basic Pilot confirmed employment eligibility for these numberholders.  In seven of these cases, the Numident records had no indication the numberholder was deceased.  Yet, other SSA records (that is, the Master Beneficiary Record) provided a date of death for the individuals.  In the other three cases, the Numident records were annotated with a transaction code of “D,” which indicates that a report of death was received by SSA, but the death was not confirmed. 

As previously mentioned in this report, we do not believe the Basic Pilot should confirm employment eligibility without some additional contact with these individuals.  Rather, we believe SSA should first reconcile its records and annotate any indications of death on the Numident—if an indicator is not already present.  Secondly, we believe the Basic Pilot should issue a tentative nonconfirmation notice to employers that more investigation is necessary for individuals with a transaction code of “D” on the Numident record (and any other code SSA chooses to use for death indicators annotated from the aforementioned reconciliation). 

Numberholders Age 90 and Over Receive No Further Screening by Basic Pilot

Within the sample of 810 noncitizens reviewed, we identified 50 (6.2 percent) individuals age 90 and over who had no indication of death on his or her Numident record or any other SSA system of record and were not currently receiving SSA benefits.  The oldest of these numberholders would be 139 in 2006.  As mentioned previously in this report, we are concerned that the Basic Pilot has no control requiring additional investigation or contact with these numberholders before confirming employment eligibility.  In fact, we tested these 50 individuals through the Basic Pilot and all were confirmed eligible for employment.  As such, we believe the Basic Pilot should be revised to issue a notice to employers that more investigation is necessary when an employee is older than a pre‑established age.

CONCLUSION

Given the scope and breadth of information held in SSA’s Numident file, we applaud the Agency on the accuracy of the data we tested.  However, we estimate that about 17.8 million Numident records contain discrepancies that may result in incorrect Basic Pilot feedback to employers.  As Congress considers legislation requiring mandatory verification of all U.S. workers’ employment eligibility through a system such as the Basic Pilot, we believe it should also consider the significant workload that may result from the millions of numberholders whose Numident records may need to be corrected.  Our review showed that the Numident records for these individuals have discrepancies in the numberholders’ names, dates of birth, citizenship status and/or death indications that would result in inaccurate feedback from the Basic Pilot. 

Because our tests included SSNs that were assigned decades ago, we recognize that some numberholders would no longer be working and would not attempt to correct their SSA and/or immigration records.  However, if even a portion of the estimated numberholders whose Numident records contained discrepancies were required to visit an SSA office to correct their information, the Agency’s workload may significantly increase until such time as the affected records were corrected. 

OTHER MATTER

DHS’ U.S. CITIZENSHIP AND IMMIGRATION SERVICES UNABLE TO LOCATE IMMIGRATION RECORDS FOR NONCITIZEN RECORDS

DHS’ U.S. Citizenship and Immigration Services (USCIS) was unable to locate immigration records for 304 (37.5 percent) of the 810 noncitizen numberholders in our sample—through its computer data match. As such, for these numberholders, we could not determine whether SSA’s Numident data matched DHS information (such as, name, date of birth and citizenship status). USCIS stated it could not match/locate these records through computer searches because of data compatibility problems. Specifically, USCIS’ Central Index System records are organized by alien number
(A-Number). Visitors or other noncitizens can be admitted under an I-94 Admission number. The records of these admissions are tracked in various DHS databases, primarily the Treasury Enforcement Communications System/Interagency Border Inspection System. We could only provide DHS with data SSA recorded on the noncitizen’s Form SS-5 or in the Numident—which generally included the numberholder’s name, date of birth, place of birth, parents’ names, approximate date of entry to the United States (date of the SSN application) and, if SSA recorded it, the noncitizen’s alien (A-number) or admission number.

We first requested immigration information from DHS for noncitizens in our sample on May 8, 2006. USCIS retained the services of a contractor, which used computer searches based on the numberholder’s name and date of birth, in attempting to locate the records in USCIS’ Central Index System. On occasion, USCIS also attempted to manually search its records for the individuals. On December 6, 2006, USCIS provided the following statement regarding its inability to match/locate some of these records.

Thank you for the opportunity to comment on those records USCIS was unable to match. The attached file you provided contains the names and information of data either not found or not matched in entirety in the Central Index System (CIS) or in manual records located here. SSA provided names and dates of birth in order for USCIS to match data and locate an A-number within USCIS' Central Index System to determine the Class of Admission. Past attempts to match SSA and USCIS data have shown that it is difficult to match records without a common numerical identifier between the two systems (Numident and CIS). Since SSA uses SSN and CIS is A-number driven it can be difficult to match records based solely on name and date of birth because slightly different names may be used with each agency, and names may change. Moreover, birth dates may be incorrect or miskeyed in one system or the other. Additionally, there may be more than one person with the same name and date of birth, making it impossible to know which record is the correct match.
It should be noted that the lack of a match does not mean the CIS system is deficient. It may well be the case that non-citizens who are not known to USCIS because they are illegal entrants have obtained SSNs, and thus have a Numident record.  Without knowing the specifics of each of the 304 cases, we cannot draw conclusions about whether the mismatch was due to inadequate information provided, or due to the fact that no USCIS record exists on the subject of the Numident record.  The possibility that some are not known to USCIS because they are not lawful entrants should not be excluded.1  USCIS Records Division has checked CIS to the best extent possible based on the names and information provided.

1 Footnote to USCIS Statement.  We would also note that a significant percentage of the records worked with overall in this study were older ones, which are the ones most likely to lack a corresponding USCIS record because they related to SSN’s issued pre-1974, when SSA did not require proof of identity or citizenship, and thus are more likely to pertain to an unlawful alien than newer Numident records.

APPENDICES

Appendix A -- Acronyms

Appendix B -- Scope and Methodology

To accomplish our objective, we performed the following steps.

• Reviewed pertinent sections of the Social Security Administration’s (SSA) policies and procedures as well as other relevant Federal laws and regulations.
• Reviewed Office of the Inspector General, Government Accountability Office and Department of Homeland Security (DHS) reports and other relevant documents.
• Randomly selected a segment of SSA’s Numident file as of November 30, 2005.   From the segment, we stratified the Social Security numbers (SSN) into three populations and randomly selected 810 Numident records for review from each:  (1) native‑born U.S. citizens, (2) foreign-born U.S. citizens and (3) non‑U.S. citizens.  Table 1 provides the actual number of SSNs in Segment 3, categorized by the aforementioned populations.

Table 1:  Details of Sample Selected for Review

• For each sample item, we obtained copies of the original Application for a Social Security Card (Form SS-5) and any subsequent SS-5s that resulted in an update to the numberholder’s Numident record (for example, a name change or change in citizenship status.)
• We then compared the Numident data (name, SSN, date of birth, and citizenship status) to the SS‑5s to ensure SSA personnel correctly recorded the information provided by the applicant. 

• We attempted to determine the legitimacy of any dates of death present on the Numident records, if applicable, by (1) researching SSA and other records to determine what State and/or jurisdiction the death occurred and (2) if located, requesting a copy of the death certificate from that location.
• For numberholders age 90 and over, who did not have a date of death annotated on their Numident records, we reviewed SSA’s Master Beneficiary Record and Supplemental Security Income Record Display to determine whether a death indication was noted on these records for the individual and whether they were currently receiving SSA benefits.
• We verified foreign-born noncitizens’ and foreign-born U.S. citizens’ “citizenship” status with DHS and/or the Department of State (State). 
• We obtained birth and, if applicable, marriage and/or death certificates for the first 50 sample items in each of the 3 populations. 
• We submitted the 1,620 U.S. citizen SSNs through the Basic Pilot, using information provided by the numberholder on his or her Form SS‑5.  Further, we submitted all noncitizen SSN records through the Basic Pilot when we could obtain DHS alien or admission numbers or when we identified any discrepancies between SSA and DHS information for the numberholder. 
• We summarized and projected our results to the entire Numident universe (all 20 segments) as of November 30, 2005.  See Appendix C for our sample results.

Underlying Assumptions
In performing our tests, we could not predict what proof of identity and employment eligibility a numberholder would present to an employer.  Therefore, we made the following assumptions.

• The information provided by the numberholder on his or her Form SS-5 was correct and would be the same data provided to an employer. 
• The information provided by DHS and/or State was correct.

Our review of internal controls was limited to obtaining an understanding of SSA’s SSN assignment process, the Numident file and the SSA/DHS Basic Pilot—and conducting the tests outlined above.  The objective of our review was to determine the reliability of the Numident file.  Accordingly, we determined that the Numident file information was generally reliable.  However, we did not test the information provided by DHS and State.  Accordingly, we cannot opine to its reliability.  Any conclusions discussed in this report, which were predicated on information provided by DHS or State, have been annotated with the appropriate qualification. 
The SSA entities audited were the Offices of Operations, Systems, and Income and Security Programs.  We conducted the audit between April and November 2006 in Birmingham, Alabama.  We conducted our audit in accordance with generally accepted government auditing standards.

Appendix C -- Sample Results

Table 1:  Sample Results and Projections for All Citizenship Categories

Table 2:  Sample Results and Projections For Individual Citizenship Strata

Table 3: Other Projections for Foreign-Born U.S. Citizens

Table 4: Other Projections for Non-U.S. Citizens

Appendix D -- Basic Pilot

The Basic Pilot is a Department of Homeland Security (DHS) initiative, supported by the Social Security Administration (SSA) through a Memorandum of Understanding between the two agencies.  The purpose of the Basic Pilot is to assist employers in verifying the employment eligibility of newly-hired employees.  The President signed The Basic Pilot Program Extension and Expansion Act of 2003 (Public Law Number 108-156) into law on December 3, 2003.  This law extended the operation of the Basic Pilot for an additional 5 years (to a total of 11 years) and expanded the operation to all 50 States not later than December 1, 2004.
As discussed with SSA and DHS staff, the Basic Pilot involves using the information in Government databases (SSA databases and, if needed, DHS databases) to determine the employment eligibility of new hires.  The Social Security number (SSN) and Alien Registration Number ("A" Number) or I-94 Number (Admission Number) are used for these checks.  The employer must complete the DHS-issued Employment Eligibility Verification Form (Form I-9) for each employee and then enter elements of this data into the Basic Pilot within 3 days of hiring, including the employee's SSN, name, date of birth, and whether the new hire indicated he or she was a U.S. citizen and, if not, the "A" Number or I-94 Number.
The system first checks the information entered against SSA's database to verify the name, SSN, and date of birth of newly-hired employees, regardless of citizenship.  When the Numident indicates United States citizenship for the newly-hired employee and the new-hire indicated that he/she is a U.S. citizen, the Basic Pilot automated system confirms employment eligibility.  If the Basic Pilot system cannot confirm employment eligibility based on the information in SSA's database or an "A" Number or I-94 Number was entered, the Basic Pilot system checks the data against DHS' database.

Appendix E -- Additional Discrepancies that Did Not Impact Basic Pilot Results

DISTRIBUTION SCHEDULE

Commissioner of Social Security                                                                                                  
Office of Management and Budget, Income Maintenance Branch                                             
Chairman and Ranking Member, Committee on Ways and Means                                          
Chief of Staff, Committee on Ways and Means                                                                           
Chairman and Ranking Minority Member, Subcommittee on Social Security                          
Majority and Minority Staff Director, Subcommittee on Social Security                                    
Chairman and Ranking Minority Member, Subcommittee on Human Resources                    
Chairman and Ranking Minority Member, Committee on Budget, House of Representatives           
Chairman and Ranking Minority Member, Committee on Government Reform and Oversight          
Chairman and Ranking Minority Member, Committee on Governmental Affairs                      
Chairman and Ranking Minority Member, Committee on Appropriations, House of Representatives         
Chairman and Ranking Minority, Subcommittee on Labor, Health and Human Services, Education and Related Agencies, Committee on Appropriations,
   House of Representatives                                                                                                           
Chairman and Ranking Minority Member, Committee on Appropriations, U.S. Senate         
Chairman and Ranking Minority Member, Subcommittee on Labor, Health and Human Services, Education and Related Agencies, Committee on Appropriations, U.S. Senate                                        
Chairman and Ranking Minority Member, Committee on Finance                                            
Chairman and Ranking Minority Member, Subcommittee on Social Security and Family Policy      
Chairman and Ranking Minority Member, Senate Special Committee on Aging                    

Social Security Advisory Board

Overview of the Office of the Inspector General

The Office of the Inspector General (OIG) is comprised of our Office of Investigations (OI), Office of Audit (OA), Office of the Chief Counsel to the Inspector General (OCCIG), and Office of Resource Management (ORM). To ensure compliance with policies and procedures, internal controls, and professional standards, we also have a comprehensive Professional Responsibility and Quality Assurance program.

Office of Audit

OA conducts and/or supervises financial and performance audits of the Social Security Administration's (SSA) programs and operations and makes recommendations to ensure program objectives are achieved effectively and efficiently. Financial audits assess whether SSA's financial statements fairly present SSA's financial position, results of operations, and cash flow. Performance audits review the economy, efficiency, and effectiveness of SSA's programs and operations. OA also conducts short-term management and program evaluations and projects on issues of concern to SSA, Congress, and the general public.

Office of Investigations

OI conducts and coordinates investigative activity related to fraud, waste, abuse, and mismanagement in SSA programs and operations. This includes wrongdoing by applicants, beneficiaries, contractors, third parties, or SSA employees performing their official duties. This office serves as OIG liaison to the Department of Justice on all matters relating to the investigations of SSA programs and personnel. OI also conducts joint investigations with other Federal, State, and local law enforcement agencies.

Office of the Chief Counsel to the Inspector General

OCCIG provides independent legal advice and counsel to the IG on various matters, including statutes, regulations, legislation, and policy directives. OCCIG also advises the IG on investigative procedures and techniques, as well as on legal implications and conclusions to be drawn from audit and investigative material. Finally, OCCIG administers the Civil Monetary Penalty program.

Office of Resource Management

ORM supports OIG by providing information resource management and systems security. ORM also coordinates OIG's budget, procurement, telecommunications, facilities, and human resources. In addition, ORM is the focal point for OIG's strategic planning function and the development and implementation of performance measures required by the Government Performance and Results Act of 1993.

-- FOOTNOTES FOLLOW --