STATE
DEPARTMENT
Pervasive Passport
Fraud Not Identified,
but Cases of
Potentially Fraudulent
and High-Risk
Issuances Are under
Review
Report to Congressional Requesters
May 2014
GAO-14-222
United States Government Accountability Office
United States Government Accountability Office
Highlights of GAO-14-222, a report to
congressional requesters
May 2014
STATE DEPARTMENT
Pervasive Passport Fraud Not Identified, but
Cases of
Potentially Fraudulent and High-Risk Issuances Are
under Review
Why GAO Did This Study
Fraudulent passports pose a significant
risk because they can be used to
conceal the true identity of the user
and potentially facilitate other crimes,
such as international terrorism and
drug trafficking. State issued over 13.5
million passports during fiscal year
2013.
GAO was asked to assess potential
fraud in State’s passport program. This
report examines select cases of
potentially fraudulent or high-risk
issuances among passports issued
during fiscal years 2009 and 2010—the
most recently available data at the time
GAO began its review. GAO matched
State’s passport data from fiscal years
2009 and 2010 for approximately 28
million issuances to databases with
information about individuals who were
deceased, incarcerated in state and
federal prison facilities, or who had an
active warrant at the time of issuance.
GAO also analyzed the passport data
to identify issuances to applicants who
provided a likely invalid SSN, which
had not been assigned at the time of
the passport application, or had been
publically disclosed. From each of
these five populations, GAO selected
nongeneralizable samples for
additional review. GAO also randomly
selected a generalizable sample from a
population of passport issuances to
applicants who used only the SSN of a
deceased individual. GAO reviewed
State’s adjudication policies, and
examined passport applications for
these populations to further assess
whether there were potentially
fraudulent or high-risk issuances. State
provided technical comments and
generally agreed with our findings. This
report contains no recommendations.
What GAO Found
Of the approximately 28 million passports issued in fiscal years 2009 and 2010
that GAO reviewed, it found issuances to applicants who used the identifying
information of deceased or incarcerated individuals, had active felony warrants,
or used an incorrect Social Security number (SSN); however, GAO did not
identify pervasive fraud in these populations. The Department of State (State)
has taken steps to improve its detection of passport applicants using identifying
information of deceased or incarcerated individuals. In addition, State modified its
process for identifying applicants with active warrants, and has expanded
measures to verify SSNs in real time. GAO referred, and State is reviewing,
matches from this analysis. The following summarizes GAO’s findings:
• Deceased individuals. As shown in the figure, GAO identified at least 1
case of potential fraud in the sample of 15 cases, as well as likely data
errors. State reviewed the cases referred by GAO, and indicated fraud could
likely be ruled out in 9 of the 15 cases; State plans to further review 6 cases.
• State prisoners. GAO found 7 cases of potential fraud among the sample of
14 state prisoner cases. State noted fraud could likely be ruled out in 10 of
the 14 cases, and intends to conduct additional reviews of 4 cases.
• Federal prisoners. None of the 15 cases in this sample had fraud indicators,
since all individuals were not actually in prison when applying for passports.
• Individuals with active warrants. GAO found five cases where State
identified the warrant and resolved it prior to issuance. As the figure shows,
GAO also identified three cases with warrants that State was not aware of or
alerted to, but should have been in State’s system for detection during
adjudication.
Summary of GAO’s Matching Analysis and Nongeneralizable Samples
a
In addition, GAO found 13,470 passport issuances to individuals who used the
SSN, but not the name, of a deceased person, as well as 24,278 issuances to
applicants who used a likely invalid SSN. GAO reviewed a 140-case
generalizable sample and a 15-case nongeneralizable sample for these two
populations, respectively, and determined the cases were likely data errors. State
has taken steps to capture correct SSN information more consistently.
Total passport issuances are solely based on the matching criteria. GAO did not verify that all
issuances from its match populations were actual fraud cases or issuances to individuals with active
warrants. Rather, it selected samples for further review and referred all matches to State.
View GAO-14-222. For more information,
contact Stephen M. Lord at (202) 512-6722 or
Page i GAO-14-222 Department of State Passport Services
Letter 1
Background 6
Examples of Potentially Fraudulent or High-Risk Passport
Issuances Found, but Pervasive Fraud Not Identified 13
Agency Comments 30
Appendix I Objectives, Scope, and Methodology 31
Appendix II Passport Application and Adjudication Process 37
Appendix III The Department of State’s (State) Use of the Social Security
Administration’s (SSA) Records for Death Checks 39
Appendix IV Description of Warrant-Matching Analysis 41
Appendix V GAO Contact and Staff Acknowledgments 43
Table
Table 1: Warrants by Type of Offense to Individuals Issued
Passports from Fiscal Years 2009 and 2010 42
Figures
Figure 1: Summary of Matching Analysis and Samples by
Population 4
Figure 2: Passport Application and Adjudication Process 8
Figure 3: Description of Matching Analysis and Sample of
Deceased Individuals 14
Figure 4: Summary of Matching Analysis and Sample of State
Prisoners 16
Figure 5: Summary of Matching Analysis and Sample of Federal
Prisoners 18
Contents
Page ii GAO-14-222 Department of State Passport Services
Figure 6: Timeline Showing When the Department of State (State)
Began Checking for Federal, State, and Local Warrants 22
Figure 7: Summary of Matching Analysis and Sample of
Applicants with Active Warrants 23
Figure 8: Summary of Whether State Identified the Warrants in
Our Sample Population 24
Figure 9: Summary of Matching Analysis and Sample of
Deceased-SSN Errors 26
Figure 10: Estimated Percentage of Causes of Incorrect SSNs
Associated with Deceased Individuals in State’s Passport
Data 28
Figure 11: Summary of Analysis and Invalid Social Security
Number Sample 29
Figure 12: Summary of Matching Analysis and Samples by
Population 35
Figure 13: Passport Application and Adjudication Process 38
Figure 14: Social Security Administration’s Responses and
Actions of Passport Specialists 40
Page iii GAO-14-222 Department of State Passport Services
Abbreviations
BOP Bureau of Prisons
CLASS Consular Lookout and Support System
DS Bureau of Diplomatic Security
EVS Enumeration Verification System
FAM Foreign Affairs Manual
FBI Federal Bureau of Investigation
FinCEN Financial Crimes Enforcement Network
FPM Fraud Prevention Manager
NCIC National Crime Information Center
SSA Social Security Administration
SSN Social Security number
State Department of State
TDIS Travel Document Issuance System
This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product may be reproduced and distributed in its entirety
without further permission from GAO. However, because this work may contain
copyrighted images or other material, permission from the copyright holder may be
necessary if you wish to reproduce this material separately.
Page 1 GAO-14-222 Department of State Passport Services
441 G St. N.W.
Washington, DC 20548
May 1, 2014
The Honorable Dianne Feinstein
Chairman
Select Committee on Intelligence
United States Senate
The Honorable Sheldon Whitehouse
Chairman
Subcommittee on Crime and Terrorism
Committee on the Judiciary
United States Senate
The Honorable Benjamin L. Cardin
United States Senate
Fraudulent passports pose a significant risk because they can be used to
conceal the true identity of the user. In addition, according to the
Department of State (State), passport and visa fraud are often committed
in connection with crimes such as international terrorism, drug trafficking,
organized crime, alien smuggling, money laundering, pedophilia, and
murder. As a result, even a few instances of passport fraud can have far-
reaching effects. A passport is an official government document that
conveys certain benefits, such as certifying an individual’s identity,
permitting a citizen to travel abroad, proving citizenship, assisting with
loan applications, and fulfilling other needs not related to international
travel. State issued over 13.5 million passport cards and books during
fiscal year 2013.
Page 2 GAO-14-222 Department of State Passport Services
Since May 2005, we have issued several reports identifying fraud
vulnerabilities in the passport issuance process.
1
In 2005 and 2007, we
reported on weaknesses in State’s information sharing with other federal
agencies, such as the Social Security Administration (SSA), as well as
opportunities to improve the agency’s oversight of passport acceptance
facilities. In 2010, we tested State’s passport issuance procedures by
using counterfeit documents and the identities of fictitious or deceased
individuals, inducing State to issue five genuine U.S. passports.
2
State
identified two of our seven applications as fraudulent during its
adjudication process; however, we were able to obtain passports using
counterfeit documents in three cases.
3
You asked that we assess potential fraud in State’s passport program.
This report examines select cases of potentially fraudulent or high-risk
issuances among passports issued during fiscal years 2009 and 2010.
We have made several
recommendations beginning in 2005 designed to help reduce passport
fraud, including that State improve information sharing with other federal
agencies, improve execution of passport fraud-detection efforts, and
strengthen internal controls at its passport-acceptance facilities. State
generally concurred with our recommendations and has taken steps to
address them.
1
GAO, State Department: Undercover Tests Show Passport Issuance Process Remains
Vulnerable to Fraud, GAO-10-922T (Washington, D.C.: July 29, 2010); State Department:
Significant Vulnerabilities in the Passport Issuance Process, GAO-09-681T (Washington,
D.C.: May 5, 2009) Addressing Significant Vulnerabilities in the Department of State’s
Passport Issuance Process, GAO-09-583R (Washington, D.C.: Apr. 13, 2009);
Department of State: Undercover Tests Reveal Significant Vulnerabilities in State’s
Passport Issuance Process, GAO-09-447 (Washington, D.C.: Mar. 13, 2009); Border
Security: Security of New Passports and Visas Enhanced, but More Needs to Be Done to
Prevent Their Fraudulent Use, GAO-07-1006 (Washington, D.C.: July 31, 2007); and
State Department: Improvements Needed to Strengthen U.S. Passport Fraud Detection
Efforts, GAO-05-477 (Washington, D.C.: May 20, 2005).
2
Suspicious identifying information and documentation included passport photos of the
same investigator on multiple applications; a 62-year-old applicant using a recently issued
Social Security number (SSN); passport and driver’s license photos showing about a 10-
year age difference; as well as the use of a California mailing address, a West Virginia
permanent address and driver’s license address, and a Washington, D.C., phone number
in the same application.
3
In the two remaining cases, State recovered the passports from the mail before they
were delivered.
Page 3 GAO-14-222 Department of State Passport Services
To examine potentially fraudulent or high-risk passport issuances in fiscal
years 2009 and 2010,
4
we matched State’s passport-issuance data for
approximately 28 million passport issuances (including passport books
and cards)
5
to databases containing information about individuals who
were (1) deceased, (2) incarcerated in a state prison facility, (3) in the
custody of the federal Bureau of Prisons (BOP), or (4) the subject of an
active warrant at the time of the passport issuance.
6
We conducted this
matching on the basis of common data elements including Social Security
number (SSN), name, and date of birth. We also analyzed the passport
data to identify issuances to applicants who provided an invalid SSN,
which was defined as an SSN that had not been assigned at the time of
the passport application, or had a high risk of misuse.
7
4
We used data from these fiscal years because they were the most recent, full fiscal years
available at the time State complied with our data request. In addition, for purposes of this
report, potentially fraudulent passport issuances are those that involved an applicant using
someone else’s identity to apply for and receive a passport. We defined high-risk passport
issuances as issuances to individuals who may pose a risk to public safety, but who did
not necessarily steal someone’s identity to apply for a passport, such as people with
active warrants for felony charges.
From each of
these five populations, we selected nongeneralizable samples for further
review. In addition, from a population of 13,470 passport issuances to
applicants who used the SSN, but not the name, of a deceased individual,
we selected a generalizable stratified random sample of 140 passport
issuances, including 70 passport issuances from both fiscal years 2009
5
According to publically available passport issuance statistics, State issued a combined
total of 28,964,775 passports during fiscal years 2009 and 2010. GAO reviewed
domestically issued passports and excluded passports issued by the Special Issuance
Agency to government travelers. We reviewed a total of 28,000,063 passport issuance
records for these fiscal years.
6
Our review included state prison data from 11 states including Alabama, Arizona,
California, Florida, Georgia, Indiana, Missouri, New York, Ohio, Texas, and Virginia. We
also obtained state prisoner data from five other states including Illinois, Louisiana,
Michigan, Pennsylvania, and North Carolina, but did not use the data from these sources
for our matching analysis for various reasons, including the absence of key fields or
delays in receiving the data. We selected these 16 states because they had the largest
prisoner populations as of December 31, 2009.
7
Prior to June 25, 2011, the Social Security Administration (SSA) issued SSNs according
to a sequential and geographic logic. SSNs that were issued after fiscal years 2009 and
2010 should not be found in passport data from that time and therefore all SSNs in the
passport data should be subject to SSA’s sequencing logic. In addition, SSNs that have
been publicly disclosed in advertisements or those used as placeholders by data entry
clerks (e.g., 012-34-5678 or 111-11-1111) are at higher risk of misuse, and may represent
a fraud indicator if found in the passport data.
Page 4 GAO-14-222 Department of State Passport Services
and 2010. We analyzed these cases to determine whether the applicant
provided the correct SSN and State recorded it incorrectly, or whether the
applicant provided the wrong SSN and State recorded the incorrect SSN
in its system. Figure 1 summarizes the focus of our matching analysis
and the related sample sizes selected for further review.
Figure 1: Summary of Matching Analysis and Samples by Population
a
These totals are solely based on the matching criteria described. We conducted additional reviews to
verify data for the sample items in the next column. We did not verify that all issuances from our
matching analysis were actual cases of fraud or issuances to individuals with active warrants. Rather,
we selected samples for additional review, and referred all matches to State for further investigation.
b
In our data matching, we used state prisoner data from 11 states including Alabama, Arizona,
California, Florida, Georgia, Indiana, Missouri, New York, Ohio, Texas, and Virginia. However, even
though we used data from all 11 states during our matching analysis, three of these state prison
databases did not have any valid matches to the passport data. In addition, data from three states did
Page 5 GAO-14-222 Department of State Passport Services
not include dates of birth, and therefore records in these databases matched to State’s passport data
based on Social Security number (SSN) and name only.
c
As described in more detail later in this report, our review indicated that some matches may be a
result of identity theft perpetrated by the state prisoner, prior to incarceration, and not the passport
applicant.
d
We selected a nongeneralizable sample of up to two prisoners incarcerated in each of the states we
reviewed for a total of 14 cases from eight different states.
e
This number includes passport issuances to people residing in halfway houses, and therefore may
not represent issuances to individuals using the identities of federal prisoners to apply for a passport.
f
These 486 passport issuances were associated with 442 unique individuals that had a total of 564
open warrants. We did not confirm that all of these warrants were associated with felony charges, but
we excluded warrants with a description of either a “traffic crime” or “misdemeanor” from our analysis.
g
As highlighted in figure 1, we selected a total of 214 passport issuances
for additional review from our five nongeneralizable and one
generalizable samples. For each of the 214 passport issuances selected,
we reviewed a copy of the original passport application, submitted the
SSN from State’s passport data to SSA for verification, and obtained
records of the passport holder’s travel activity from the Financial Crimes
Enforcement Network (FinCEN), a bureau of the U.S. Department of the
Treasury. We also reviewed State’s documentation of additional
investigative steps the agency took, if any, to resolve fraud indicators
during the adjudication of the passports. Where applicable, we obtained
additional documentation about the death, incarceration, or fugitive status
of applicants from federal and state agencies. For this review, we
included only issued passports; we did not examine passport applications
that were rejected by State or abandoned by the applicant. Furthermore,
we did not attempt to identify all possible types of passport fraud.
The generalizable stratified random sample of 140 passport issuances included 70 passport
issuances from both fiscal years 2009 and 2010.
We assessed the reliability of State’s passport data, TECS travel-activity
data provided by FinCEN, SSA’s full death file, state and federal prisoner
data, and data on individuals with open warrants provided by the U.S.
Marshals Service by reviewing relevant documentation, interviewing
knowledgeable agency officials, and examining the data for obvious
errors and inconsistencies.
8
8
TECS is a data repository to support, among other things, law enforcement “lookouts”
and border screening. TECS is owned and managed by the U.S. Customs and Border
Protection with the Department of Homeland Security.
With the exception of prisoner data from five
states, which we did not use, we concluded that all of the data we used
were sufficiently reliable for the purposes of this report. We examined
State’s policies, guidance, including the Foreign Affairs Manual (FAM),
Page 6 GAO-14-222 Department of State Passport Services
and other materials provided to passport specialists. We reviewed
changes to State’s controls since fiscal years 2009 and 2010 with respect
to preventing certain fraudulent or high-risk passport issuances. We also
interviewed State officials, observed the adjudication process at a
passport facility, and reviewed the 214 passport applications in our
samples to further assess whether there were potentially fraudulent or
high risk issuances. For a more-detailed description of our objectives,
scope, and methodology, see appendix I.
We performed this audit from March 2010 through May 2014 in
accordance with generally accepted government auditing standards.
9
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objective. We believe that
the evidence obtained provides a reasonable basis for our findings and
conclusions based on our audit objective.
According to agency officials and guidance posted on State’s public
website, applicants can apply for a U.S. passport in one of three ways: in
person at an acceptance facility, by mail (for renewal applications), or at a
passport facility that offers acceptance services (typically expedited
applications). Applicants submit documents, such as a birth certificate or
driver’s license, to passport acceptance agents to provide evidence of
citizenship, or noncitizen nationality, and proof of identity. The acceptance
agents are to watch the applicant sign the application, review submitted
documents for completeness, and check for application inconsistencies.
For example, acceptance agents are to assess whether photographs and
descriptions in the identification documents match the applicant. If an
acceptance agent suspects that an applicant has submitted fraudulent
information or exhibits nervous behavior, the acceptance agent is
instructed to accept the application and complete a checklist indicating
9
The extended period required for our review was a result of various factors, including
data-sharing negotiations with State, the time required to receive and review requested
documentation, extensive data preparation and analysis involving multiple agency
databases, and State’s requirement to review all sensitive information on-site at the State
Department.
Background
Passport Application
Process
Page 7 GAO-14-222 Department of State Passport Services
the reason for suspected fraud. The agents are to then send the
application, checklist, and photocopy of the identification to State’s Fraud
Prevention Manager (FPM). Acceptance agents are not State employees;
however, State provides training, as well as detailed guidance that
governs their work. State also conducts periodic inspections and audits of
acceptance facilities to ensure compliance with regulations and policies.
According to State officials, the most common way to renew a passport is
by mail. An individual with a passport issued during the previous 15 years
may renew it by submitting a mail-in application, along with the previously
issued passport, a recent photograph, and documentation of a name
change, if applicable. Applications submitted by mail or at an acceptance
facility are sent to a Department of the Treasury contracted lockbox
service provider for data entry and payment processing. The lockbox
service provider converts handwritten or typed text into electronic data
and deposits passport fees paid by the applicant. Once the lockbox data
entry and payment are complete, the electronic data and paper passport
application are sent to passport-issuing facilities around the United States
for adjudication.
Applicants who demonstrate a need for in-person expedited service for
either a first-time issuance or a renewal may submit their applications
directly to a passport-issuing facility. State employees at these facilities
accept passport fees and enter application data directly into State’s
electronic processing system, called the Travel Document Issuance
System (TDIS), before forwarding the application for expedited
adjudication.
Figure 2 provides an overview of the passport application and
adjudication process for applications received in person at an acceptance
facility, by mail, or at a passport facility that offers acceptance services
(see app. 2 for static version of this figure).
Page 8
GAO-14-222 Department of State Passport Services
Figure 2: Passport Application and Adjudication Process
Interactive Graphic
Print Version: Click here or go to appendix II.
Application submitted
at a passport facility
Application mailed
Application submitted
at an acceptance facility
Instructions: Roll over the for more information.
Passport specialists adjudicate all applications
Data entry and payment processing
Passport specialist checks
identity and citizenship
documentation
Role of Passport Specialist
Passport specialist reviews
results of database checks
Passport specialist refers applications
with fraud or ineligibility indicators to
supervisor or authorizing entity
Passport specialist addresses issues
identified as a result of database
checks and review of documentation
Authorizing entity
reviews application
Issuance approved
Issuance denied
X
Source: GAO review of the Department of State (State) data. Art Explosion (images).
Page 9 GAO-14-222 Department of State Passport Services
As we have noted in previous products, each passport application is to be
individually reviewed by a passport specialist during a process known as
adjudication.
10
State’s FAM specifies the steps passport specialists must
take to address various fraud indicators. According to State documents,
specialists are responsible for reviewing applications and documents
establishing the applicants’ identity and citizenship, as well as conducting
various checks, as described below. Depending on the results of the
adjudication, passport specialists may approve or deny the passport
issuance, conduct additional checks, request more information from the
applicant, or forward the application for additional review by their
supervisor, the FPM, or by offices in Passport Headquarters.
11
Once a
passport has been issued, the application is scanned and archived.
Passports issued to individuals 16 years or older are generally valid for 10
years.
12
Several federal statutes and regulations either require or permit State to
withhold a passport from an applicant in certain situations. For example,
State must withhold passports from individuals who are in default on
certain U.S. loans, who are in arrears of child support in an amount
determined by statute, or who are imprisoned, on parole, or on
supervised release as a result of certain types of violations of the
Controlled Substances Act, Bank Secrecy Act, and some state-level drug
laws.
13
Likewise, State may choose to refuse a passport to applicants
who are the subject of an outstanding local, state, or federal warrant of
arrest for a felony, or the subject of probation conditions or criminal court
orders that forbid the applicant from leaving the country and the violation
of which could result in the issuance of a federal arrest warrant.
14
During the adjudication process, passport specialists are to review
applications and results of checks against various databases to detect
fraud and suspicious activity, and for other purposes. Application data are
10
For example, see GAO-10-922T.
11
According to a State official, Passport Headquarters consists of eight offices, such as
the Office of Adjudication and the Office of Technical Operations.
12
A passport issued to an applicant who is under 16 years old is generally valid for 5
years.
13
22 U.S.C. § 2671(d), 42 U.S.C. § 652(k), and 22 U.S.C. § 2714.
14
22 C.F.R. § 51.60(b).
Passport Adjudication
Process
Page 10 GAO-14-222 Department of State Passport Services
entered into TDIS, State’s electronic processing system. TDIS
automatically checks applicants’ names against a number of sources,
including SSA’s death records and a database of warrants. For example,
TDIS automatically checks key identifying information of all passport
applicants against SSA’s full death file, as well as a database of felony
warrants for certain crimes. Passport specialists are to compare the
application to the information in TDIS to make sure it was entered
properly and to identify missing information.
Passport specialists are also to review the results of automatic checks
during a process State refers to as “the front-end” process of adjudication.
For instance, during this process, passport specialists are to determine
whether an applicant currently holds a passport, has a history of lost or
stolen passports, or has already submitted a passport application.
According to State officials, such checks are intended to facilitate the
identification of suspicious activity and prevent multiple passport
issuances to the same person. Passport specialists also are to consider
the results from facial recognition technology which is used to help
prevent the issuance of passports to individuals using false identities and
people who should be denied passports for other legal reasons, such as
terrorists in the Federal Bureau of Investigation’s (FBI) terrorist
database.
15
In addition, passport specialists may employ commercial
databases and other tools during the adjudication process to assist in
confirming an applicant’s identity or citizenship.
15
In technical comments, State officials clarified that specialists do not deny passport
applications based solely on the results of facial recognition technology. According to
officials, facial recognition technology is one of many tools specialist use to determine
whether an application should be referred for further review and investigation. Officials
added that State does not deny passports to individuals identified in the FBI’s terrorist
database without additional review and investigation by the appropriate office.
Page 11 GAO-14-222 Department of State Passport Services
In April 2007, State and SSA signed an information-exchange agreement
that allows State to query SSA’s records for verifying applicants’ identities
and identifying deceased individuals. In accordance with this agreement,
State’s TDIS automatically queries SSA’s Enumeration Verification
System (EVS) to verify that a passport applicant’s SSN, name, and date
of birth match the records at SSA.
16
EVS includes a death indicator based
on SSA’s full death file of approximately 98 million records, which aids
State in identifying applicants using the identity of a deceased individual
to apply for a passport.
17
16
26 U.S.C. § 6039E requires passport applicants to provide an SSN, if they have one,
when applying for or renewing a passport. However, passports may be issued to
applicants who do not have an SSN.
In most cases, State’s controls will not flag an
applicant as deceased unless certain fields such as the SSN, name, and
date of birth all match the identifying information of a deceased individual.
According to State’s procedures, passport specialists must refer any
applications with a positive death indicator to State’s FPM for additional
review, since the match may indicate a case of stolen identity. The FPM
reviews all applications referred to it by passport specialists to determine
whether the identifying information on the passport application is in fact
associated with a deceased individual. The FPM can approve the
passport application once it has reviewed and resolved any indicators of
17
As GAO previously reported in May 2013, (GAO, Social Security Administration:
Preliminary Observations on the Death Master File, GAO-13-574T [Washington, D.C.:
May 8, 2013]), the Social Security Act places limitations on SSA’s sharing of state-
reported death information; SSA removes the state-reported records from the full death
file and provides the public death file, or public Death Master File, to the Department of
Commerce’s National Technical Information Service, which sells it through a subscription
service. Since October 2009, State used its subscription to the public death file, which
excludes state-reported death information and is available publicly to any interested party
for a fee. State officials noted in their technical comments that all applications are now
checked using SSA’s real-time verification system, and that State uses the public death
file in exceptional circumstances, such as when the real-time system is unavailable or for
postissuance audits. Unlike the public death file, the full death file contains all death
records, including state-reported death information, and is available to federal benefit-
paying agencies; however, State has access to certain data elements from the full death
file as a result of its information-exchange agreement with SSA.
Selected Controls for
Detecting Potentially
Fraudulent and High-Risk
Issuances
SSN Verification and Data
Checks for Deceased
Individuals
Page 12 GAO-14-222 Department of State Passport Services
potential fraud. See appendix III for additional details on State’s use of
SSA’s records for death checks.
In fiscal years 2009 and 2010, the years of passport issuances we
reviewed, State did not have access to federal and state prisoner
databases in order to check whether applicants’ identities matched those
of incarcerated individuals. Since then, State has taken steps to explore
access to such databases. For example, in June 2013, State entered into
a data-sharing agreement with the BOP in order to access federal
prisoner data. In addition, officials told us that in December 2013, State
completed the first phase of a pilot project using prisoner data from two
states, Florida and Rhode Island, to identify whether applicants are
fraudulently using identities of state prisoners. We provide additional
details on State’s initiatives to improve data checks for incarcerated
individuals in a subsequent section.
In 2002, the Marshals Service began transmitting certain warrant data to
State for use during the passport adjudication process. Since then, the
information State receives has changed to include additional warrants
from the FBI, as described in detail below. To help State determine
whether an applicant may have an active warrant for a felony charge,
TDIS automatically checks applicants’ identifying information in State’s
Consular Lookout and Support System (CLASS), a database that
maintains warrant data.
18
18
In addition to warrant data, CLASS contains other information, such as data from the
FBI’s Terrorist Screening Center database and information from the Department of Health
and Human Services about individuals delinquent on child support.
TDIS indicates a possible match if certain data
elements from the passport application, such as the name, SSN, date of
birth, place of birth, or gender, matches information in CLASS within
certain parameters. State’s policies require that passport specialists refer
likely matches in CLASS to State’s passport legal office. Officials said
paralegals in the passport legal office are to review the information and
contact the warrant issuer to confirm the identity of the subject in the
warrant against the passport applicant, verify that the warrant is active
and related to a felony charge, and further coordinate, as necessary. The
passport legal office may also use commercial databases, or photographs
obtained from the warrant issuer, to confirm applicants’ identities. In
technical comments, State officials clarified that the passport legal office
is authorized to deny the passport issuance when it determines, or is
informed by a competent authority, that the applicant is the subject of an
Data Checks for Incarcerated
Individuals
Data Checks for Individuals
with Active Warrants
Page 13 GAO-14-222 Department of State Passport Services
outstanding federal, state, or local warrant of arrest for a felony crime. In
addition, the passport legal office can authorize the passport issuance if it
determines, upon additional review, that there was not in fact a legitimate
match in CLASS. The legal office may approve an issuance in cases
where the warrant was closed, associated with a misdemeanor charge, or
for other reasons, such as a request by law enforcement agencies.
Of the combined total of approximately 28 million passport issuances we
reviewed from fiscal years 2009 and 2010, we found instances of
issuances to individuals who applied for passports using identifying
information of deceased or incarcerated individuals, as well as applicants
with active felony warrants. The total number of cases we identified
represented a small percentage of all issuances during the two fiscal
years, indicating that fraudulent or high-risk issuances were not
pervasive. We also determined that State’s data contained inaccurate
SSN information for thousands of passport recipients. Most of the
instances in which there was inaccurate SSN information appeared to be
applicant or State data-entry errors, rather than fraud. Since fiscal years
2009 and 2010, State has taken steps to improve its detection of passport
applicants using the identifying information of deceased or incarcerated
individuals. In addition, State modified its process for identifying
applicants with active warrants, and has expanded measures to verify
SSNs in real time.
Out of a combined total of approximately 28 million passport issuances
we reviewed from fiscal years 2009 and 2010, we identified 181
passports issued to individuals whose name and SSN both appeared in
SSA’s full death file, suggesting that the applicant may have
inappropriately used the identity of a deceased person.
19
19
These 181 passport issuances were associated with 167 unique individuals.
To ensure that
Examples of
Potentially Fraudulent
or High-Risk Passport
Issuances Found, but
Pervasive Fraud Not
Identified
Passport Issuances to
Applicants Associated with
Deceased Individuals,
Prisoners, and Individuals
with Active Warrants
Issuances to Applicants Using
Identifying Information of
Deceased Individuals
Page 14 GAO-14-222 Department of State Passport Services
our matches did not contain legitimate applicants who died shortly after
submitting their applications, we included only individuals who had died
more than 120 days before the passport issuance.
20
Figure 3: Description of Matching Analysis and Sample of Deceased Individuals
Figure 3 summarizes
our matching analysis and sample results.
a
It is not possible to determine from data matching alone whether the
passport issuance was appropriate or fraudulent without reviewing the
facts and circumstances for each individual case from the 181 passport
issuances. Thus, we randomly selected a nongeneralizable sample of 15
cases for additional analysis. For each case, we attempted to verify death
information from SSA’s full death file by obtaining a copy of the death
certificate and confirming that SSA’s most-current records listed the
individual as deceased. We also requested TECS travel data from
FinCEN and reviewed open-source information to search for additional
fraud indicators. The following information provides additional details on
the 15 cases.
These totals are solely based on the matching criteria described. We conducted additional reviews to
verify data for the sample items in the next column. We did not verify that all issuances from our
matching analysis were actual cases of fraud or issuances to individuals with active warrants. Rather,
we selected samples for additional review, and referred all matches to State for further investigation.
• In one case, the applicant applied for and received an expedited
passport by mail in January 2009 using the SSN, name, and date of
birth of a deceased individual. The SSA’s full death file and the death
certificate indicated that the purported applicant had died in May
20
We selected 120 days after death to allow for approximately 60 days of passport
application processing time and 60 days of lag time in reporting an individual’s death to
SSA for inclusion in the full death file.
Page 15 GAO-14-222 Department of State Passport Services
2008. According to TECS travel data, the passport was used in June
2009 to fly to the United States from Mexico and had not been used
again as of June 2013. As a result of information we provided, State
reviewed this case in 2013 and determined that the applicant
appeared to be an imposter. State officials noted that the application
should have been referred to the FPM during adjudication, because it
contained multiple fraud indicators. State officials said this case
should be referred to the Bureau of Diplomatic Security (DS) for
further investigation.
• In another case, the applicant’s passport issuance was delayed by
more than a year because her name mistakenly appeared in SSA’s
full death file. In our May 2013 testimony, we found that SSA’s data
contained a small number of inaccurate records, and SSA has stated,
in rare instances, it is possible for the records of a person who is not
deceased to be included erroneously in the death file.
21
• In 4 of the 15 cases, the applicant used a similar name to, as well as
the same SSN as, a deceased individual. For each of the four cases,
we verified the death information in SSA’s full death file by obtaining a
copy of the deceased person’s death certificate. However, State
officials said fraud could likely be ruled out in all four cases for various
reasons, such as the inadvertent use of an incorrect SSN.
Situations
where a living individual is inappropriately listed as deceased in SSA’s
records can create a hardship for the person who has been falsely
identified as deceased. This case highlights one of the challenges
State encounters when querying SSA’s full death file, and illustrates
why State reviews applicants with death indicators on a case-by-case
basis.
• In 9 of the 15 cases, we could not verify the death of the applicants
because we were unable to identify the state in which the individual’s
death was recorded (possibly because the applicant was not
deceased) or because state officials would not or could not provide
the death certificate to us. State’s subsequent review of these cases
indicated that fraud could likely be ruled out in four cases, and that
five of the cases should be referred to DS for further investigation.
As of May 2014, we have referred all 181 passport issuances we
identified from our matching analysis using SSA’s full death file, including
21
GAO-13-574T.
Page 16 GAO-14-222 Department of State Passport Services
the 15 cases we examined in more detail, to State for further review and
investigation.
Out of the combined total of approximately 28 million passport issuances
we reviewed from fiscal years 2009 and 2010, we identified 68 issuances
to individuals who used an SSN, name, and in some cases, date of birth
of a state prisoner on their passport application.
22
Without reviewing the
facts and circumstances for each case, it is not possible on the basis of
data matching alone to determine the extent to which these instances
represent fraudulent issuances. Thus, from the group of individuals
related to the 68 issuances, we selected 14 cases for further review.
23
Figure 4: Summary of Matching Analysis and Sample of State Prisoners
For
each sample case, we obtained additional documentation from state
departments of corrections to verify key data fields for these passport
recipients. Figure 4 summarizes our matching analysis and sample
results.
a
22
In our data matching, we used state prisoner data from 11 states including Alabama,
Arizona, California, Florida, Georgia, Indiana, Missouri, New York, Ohio, Texas, and
Virginia. Three of these state prison databases did not have any valid matches to the
passport data. In addition, data from three states with matches did not include dates of
birth, and therefore records in these databases matched to State’s passport data based
on SSN and name only.
These totals are solely based on the matching criteria described. We conducted additional reviews to
verify data for the sample items in the next column. We did not verify that all issuances from our
matching analysis were actual cases of fraud or issuances to individuals with active warrants. Rather,
we selected samples for additional review, and referred all matches to State for further investigation.
23
The 68 passport issuances were associated with 61 unique individuals. In addition, our
review indicated that some matches may reflect situations in which a prisoner stole the
applicant’s identity prior to incarceration, resulting in matches to State’s passport data that
do not represent fraud.
Issuances to Applicants Using
the Identifying Information of
State and Federal Prisoners
Page 17 GAO-14-222 Department of State Passport Services
b
Our review indicated that some matches may be a result of identity theft perpetrated by the state
prisoner, prior to incarceration, and not by the passport applicant. In addition, data from three states
with matches did not include dates of birth, and therefore records in these databases matched to
State’s passport data based on SSN and name only.
c
From our nongeneralizable sample of 14 cases, we identified seven
passport applicants who may have fraudulently used the identities of state
prisoners, since the incarcerated individuals could not have physically
appeared at a passport facility to submit their applications.
We selected a nongeneralizable sample of up to two prisoners incarcerated in each of the states we
reviewed for a total of 14 cases from eight different states.
24
We could not conclusively determine that all our sample cases or
matches represented passport fraud, because for instance, it is possible
that the state prisoner may have stolen the identity of the applicant prior
to incarceration. For example, we identified two cases involving data from
the same prison facility in which the prisoner had an alias name, in
addition to an SSN and date of birth, that matched the information of the
passport applicant. We provided information on all our matches, including
the 14 state prisoner cases in our sample, to State for review. According
to officials, State’s review of these cases included, but was not limited to,
an assessment of fraud indicators in the passport applications, and
review of the applicants’ information in commercial and internal
databases. State determined that fraud could likely be ruled out in eight
cases. Officials initially said they should refer the remaining six cases to
DS for further investigation. In their technical comments on a draft of this
report, State officials said they conducted a second review of the six
remaining cases and determined that two individuals used their true
identities on their passport applications, and they ultimately referred four
cases to DS for investigation.
The seven
remaining cases in our state prisoner sample of 14 individuals were either
not incarcerated at the time of application submission, applied for
passports using mail-in applications, or represented possible identity theft
by the prisoner prior to incarceration. Federal regulations do not prohibit
State from issuing passports to prisoners; however, according to officials,
State’s policy is to deny passport issuances to individuals who are
incarcerated at the time of application submission.
24
These seven individuals applied for a passport using a DS-11 application, which must
be submitted in person at an acceptance facility or passport agency.
Page 18 GAO-14-222 Department of State Passport Services
Of the four cases in our state prisoner sample that State officials referred
to DS, we identified three instances where the passport was used to
cross an international border during the prisoners’ periods of
incarceration. These cases highlighted the active use of passports
obtained by potentially fraudulent means. We verified this travel activity
by comparing the names, dates of birth, and passport numbers in State’s
passport data for these cases with TECS travel data provided by FinCEN.
The TECS travel log for the three cases showed that the individuals used
the passports for international travel at least once during the prisoners’
periods of incarceration. In one case, an individual used the passport
obtained by potentially fraudulent means to cross the U.S.-Mexico border
more than 300 times.
In addition to our analysis of state prisoners, we also identified 206
passport issuances to individuals who used an SSN, name, and date of
birth in their applications that matched identifying information in the BOP’s
federal prisoner data. However, the data we received included individuals
residing in halfway houses. Unless otherwise stated in the conditions of
release for parole, passport issuances to individuals living in halfway
houses are legally permissible. Since we focused our in-depth analysis on
a nongeneralizable sample of 15 cases, we did not determine the extent
to which the 206 cases represented individuals in federal prison facilities
as opposed to halfway houses. Figure 5 summarizes our matching
analysis and sample of 15 cases.
Figure 5: Summary of Matching Analysis and Sample of Federal Prisoners
a
These totals are solely based on the matching criteria described. We conducted additional reviews to
verify data for the sample items in the next column. We did not verify that all issuances from our
matching analysis were actual cases of fraud or issuances to individuals with active warrants. Rather,
we selected samples for additional review, and referred all matches to State for further investigation.
b
This number includes passport issuances to people residing in halfway houses, and therefore may
not represent issuances to individuals using the identities of federal prisoners to apply for a passport.
Page 19 GAO-14-222 Department of State Passport Services
From our sample of 15 cases, we did not identify any individuals who
applied for a passport using the identity of a federal prisoner in their
passport application. We determined that at least 9 of the 15 individuals
were living in BOP halfway houses when the passport application was
submitted. Moreover, we did not find any indications of identity theft. The
other six individuals were either not in a halfway house at the time of
application submission, or we were unable to determine, on the basis of
the information provided, their location after they were released from a
federal prison facility.
25
In fiscal years 2009 and 2010, officials said State did not have access to
federal and state prisoner databases in order to check whether
applicants’ identities matched those of incarcerated individuals. In June
2013, State entered into a data-sharing agreement with the BOP that will
allow it to access federal prisoner data, including information about
individuals incarcerated in federal facilities or halfway houses. In addition,
State obtained data-sharing agreements with two individual state
departments of corrections, Florida and Rhode Island, as part of a pilot
project to identify whether applicants fraudulently used the identities of
state prisoners. State officials said these states represent different
geographical regions and a large and small inmate population, and both
had technical capabilities to transfer data efficiently and securely to State
for adjudication purposes. Officials said in their technical comments that
State completed the first phase of the pilot project in December 2013.
This phase included the development of search criteria for detecting the
fraudulent use of prisoners’ identities. According to officials, State referred
three potential fraud cases to DS for further investigation as a result of
this effort. Officials also reported in their technical comments that State
plans to acquire prisoner data from other states, and that it is developing
best practices for obtaining such data. In addition, officials noted that
State is in the early stages of planning a second phase of the pilot project.
However, the documentation for these six
individuals indicated that they were not incarcerated when the application
was submitted.
State officials highlighted various challenges with respect to using
prisoner data during adjudication, including technical requirements and
issues related to data transmission, as well as potential legal limitations.
25
Of these six individuals, one was incarcerated in a federal prison facility at the time of
passport issuance. However, the individual applied for the passport prior to being in
federal custody, and was subsequently issued a passport once his sentence began.
Page 20 GAO-14-222 Department of State Passport Services
For example, according to BOP officials, State and the BOP will have to
develop a technical infrastructure to facilitate sharing of federal prisoner
data, which officials expected to occur no later than the end of fiscal year
2014. Similarly, with respect to state prisoner data, State officials noted
that data from state departments of corrections would need to be
automatically transmitted to allow for the updating of information on a
consistent basis. In technical comments, State officials clarified that they
would prefer to receive data from individual state departments of
corrections on a real-time basis; however, the frequency with which State
receives these data is not a factor in determining whether State enters
into a data-sharing agreement with a department of corrections. State
officials also highlighted issues regarding the compatibility of systems
from various states, as well as concerns about poor data that could lead
to false matches and delays in processing passport applications.
Moreover, State officials told us legal limitations may prevent the transfer
of state-level inmate data; however, State did not report having such
challenges working with Florida and Rhode Island during its pilot project.
Out of a combined total of approximately 28 million passport issuances
we reviewed from fiscal years 2009 and 2010, we identified 486
issuances to individuals using the SSN, name or alias, and date of birth of
people with active warrants on their passport applications.
26
We could not
determine from matching analysis alone whether all warrants were
associated with felonies, but our analysis excluded warrants with a
description of either a “traffic crime” or “misdemeanor” in data provided by
the Marshals Service (see app. IV for additional details). The type of
warrant data State has received for detecting active felony warrants
through CLASS has changed over time. In 2002, the Marshals Service
began providing State with certain warrant information in CLASS. These
only included Class 1 warrants, which is a designation for warrants the
Marshals Service enters and maintains in the National Crime Information
Center (NCIC) database, a criminal database that provides the warrant
data to CLASS.
27
26
These issuances include individuals whose SSN, partial name (or alias), and date of
birth in warrant data provided by the Marshals Service matched information in State’s
passport data. In addition, these 486 passport issuances were associated with 442 unique
individuals with a total of 564 open warrants.
According to State officials, the FBI began providing
State with federal felony warrants in 2005 for use during the adjudication
27
The NCIC database is an electronic repository of data on crimes and criminals of
nationwide interest and a locator file for missing and unidentified persons.
Issuances to Individuals with
Active Warrants
Page 21 GAO-14-222 Department of State Passport Services
process. In late July 2009, officials said State began receiving state and
local warrants from the FBI for crimes of varying degrees of severity,
including misdemeanors, serious felonies, and nonserious felonies.
28
28
FBI officials told us the bureau completed modifications of its warrant data for State in
August 2007 to include not only federal warrants, but also state and local warrants.
According to State officials, it would have been infeasible for State to receive or use state
or local warrant data in 2007, given the technical and operational requirements for
preparing CLASS, guidance to passport specialists, and instructions to offices that assist
in the passport adjudication process. Officials added that State’s regulatory authority to
deny or revoke passports on the basis of an outstanding state or local felony warrant was
not effective until February 2008. Moreover, the Memorandum of Understanding between
the FBI and State for sharing state and local warrants was not signed by both parties until
January 2009.
According to State officials, the high volume of warrant cases was
unmanageable and State had no authority to take action on misdemeanor
warrants. Thus, in November 2010, State officials said they updated
CLASS so that it included only state or local warrants connected to more-
serious felonies they selected. Officials also said CLASS is updated daily
with information provided by the Marshals Service and the FBI, and
currently contains information for federal, state, and local felony warrants
related to State’s selected felony charges. Figure 6 illustrates the
evolution in State’s data checks for warrants.
Page 22 GAO-14-222 Department of State Passport Services
Figure 6: Timeline Showing When the Department of State (State) Began Checking for Federal, State, and Local Warrants
a
“Class 1” is a designation of the Marshals Service for warrants it enters and is responsible for in the
NCIC database, a criminal database that sources the warrant data in Consular Lookout and Support
System (CLASS).
b
From the population of 486 issuances with active warrants that we
identified through matching, we randomly selected a nongeneralizable
sample of 15 individuals for additional analysis. Figure 7 summarizes our
matching analysis and sample results.
According to officials, in November 2010, State selected certain felonies for inclusion in State’s
CLASS in response to a high volume of state and local warrants it received in July 2009, which also
included warrants for misdemeanors.
Page 23 GAO-14-222 Department of State Passport Services
Figure 7: Summary of Matching Analysis and Sample of Applicants with Active Warrants
a
These totals are solely based on the matching criteria described. We conducted additional reviews to
verify data for the sample items in the next column. We did not verify that all issuances from our
matching analysis were actual cases of fraud or issuances to individuals with active warrants. Rather,
we selected samples for additional review, and referred all matches to State for further investigation.
b
According to the Marshals Service, all 15 of the individuals in our
nongeneralizable sample had warrants related to felony charges, 3 of
which the Marshals Service was responsible for executing. Fugitives with
felony warrants may pose a risk to public safety, and passports could help
them evade capture by law enforcement agencies. State may choose to
refuse a passport to applicants who are the subject of an outstanding
felony warrant. In our analysis of the 15 cases, we took into account the
evolution in State’s controls. Figure 8 summarizes our review of the cases
in our sample.
These 486 passport issuances were associated with 442 unique individuals with a total of 564 open
warrants. Beyond our sample items, we did not confirm whether the warrant was associated with a
felony charge. However, these records excluded all warrants with a description of either a “traffic
crime” or “misdemeanor.”
Page 24 GAO-14-222 Department of State Passport Services
Figure 8: Summary of Whether State Identified the Warrants in Our Sample Population
a
According to officials, State received a notice from the relevant U.S. District Court regarding this
individual that indicated the court order for the person had ceased and the individual was sentenced
to probation. As a result, State officials said the warrant for this individual was not included in the
Consular Lookout and Support System (CLASS) for State’s warrant check.
b
Upon further review of our original sample of 15 cases, we identified one individual who had a state
or local warrant that was not issued until after the applicant applied for his passport. In addition, our
review of court documents for another case indicated the individual was unlikely to have been a
fugitive when applying for the passport, since he was arrested and faced criminal charges after the
warrant was issued, but about a decade before submitting the application. As a result, we did not
further review these cases as part of our sample.
Page 25 GAO-14-222 Department of State Passport Services
Among the 13 applicants we reviewed in detail, we found five cases with
warrants that State identified during the adjudication process. For
instance, after detecting the warrant in one case, State’s passport legal
office ultimately determined the passport applicant was a victim of identity
theft and was not the subject of the warrant. In another case, State
identified a warrant at the state or local level for an applicant who applied
for a passport at a time when State’s controls had begun checking for
such warrants. In this case, State’s passport legal office authorized the
passport issuance to the individual after concluding the associated charge
was for a misdemeanor crime, as opposed to a felony offense. We also
identified five cases where the applicants had outstanding state or local
felony warrants on the application date, and State’s CLASS did not have
data for such warrants when the individuals applied.
29
In the other 3 of the
13 sample cases we reviewed, we found no indications that State was
aware of or alerted to the individuals’ warrants at the time they applied for
passports, even though it appeared that CLASS should have included the
warrant data.
30
We referred all passport issuances we identified from our
matching analysis, including our sample cases, to State for further review
and investigation.
29
For one of these cases, State provided additional details on a subsequent passport
issuance. Specifically, State highlighted a case of passport fraud involving an individual in
our sample that occurred after the period of passport issuances we reviewed from fiscal
years 2009 and 2010.
30
Officials of the Marshals Service told us there can be delays between the date a warrant
is issued and the date a law enforcement agency validates it in the NCIC database.
Alternatively, the officials said, law enforcement agencies may elect not to enter the
warrant into a federal database at all. In such circumstances, State would not be alerted to
the warrant, regardless of whether it was federal, state, or local, because CLASS would
not contain information about it for use during the passport adjudication process.
Page 26 GAO-14-222 Department of State Passport Services
Out of the combined total of approximately 28 million passport issuances
we reviewed from fiscal years 2009 and 2010, we found 13,470 passport
issuances to individuals who submitted an SSN associated with a
deceased individual,
31
but where the name used in the passport
application did not match the name of the deceased individual.
32
Figure 9: Summary of Matching Analysis and Sample of Deceased-SSN Errors
As we
previously noted, we analyzed these cases to determine whether the
applicant provided the correct SSN and State recorded it incorrectly, or
whether the applicant provided the wrong SSN and State recorded the
incorrect SSN its system. Specifically, from this population, we selected a
stratified random sample that consisted of 140 passport issuances,
evenly divided between fiscal years 2009 and 2010 (see fig. 9). We refer
to these cases below as deceased-SSN errors.
a
These totals are solely based on the matching criteria described. We conducted additional reviews to
verify data for the sample items in the next column. We did not verify that all issuances from our
matching analysis were actual cases of fraud or issuances to individuals with active warrants. Rather,
we selected samples for additional review, and referred all matches to State for further investigation.
b
The generalizable stratified random sample of 140 passport issuances included 70 passport
issuances from both fiscal years 2009 and 2010.
31
State may issue multiple passports to the same individual, such as when the applicant
applies for both a passport book and a passport card. These 13,470 passport issuances
were associated with 12,781 unique individuals.
32
We also found 181 passport issuances where both the applicant’s name and SSN
matched the SSN and name of a deceased individual. These issuances are described in
detail in the previous section under the subheading “Issuances to Applicants Using
Identifying Information of Deceased Individuals.”
State’s Records Contained
Erroneous SSNs for
Thousands of Passport
Issuances as a Result of
Applicant or State Data-
Entry Error
Page 27 GAO-14-222 Department of State Passport Services
We estimated that approximately 50 percent of the 13,470 cases with
deceased-SSN errors were instances where the applicant provided an
SSN that did not belong to him or her.
33
33
From 13,470 passport records with names that did not match the corresponding record
in the SSA full death file, we selected a stratified random sample of 140 passport
issuances. All estimates from this sample have a margin of error of +/-9 percentage points
or fewer unless otherwise noted. See app. I for more details.
We did not identify any other
evidence of potential fraud in this group of cases, which suggests that
applicants may have made mistakes in filling out the passport application.
We estimated that in approximately 44 percent of the 13,470 cases with
deceased-SSN errors applicants provided the correct SSNs, but State
entered them incorrectly into TDIS. State officials could not provide an
explanation for the errors related to these cases, which included
applications with handwritten SSNs that were difficult to read, as well as
typed SSNs. The remaining 6 percent of cases did not fall into either of
these categories. Such cases included instances where the applicant did
not provide an SSN, or where we were unable to ascertain the applicant’s
actual SSN. Figure 10 summarizes the issuances we reviewed involving
deceased-SSN errors.
Page 28 GAO-14-222 Department of State Passport Services
Figure 10: Estimated Percentage of Causes of Incorrect SSNs Associated with
Deceased Individuals in State’s Passport Data
Notes: All percentage estimates in figure 10 are generalized to the population of 13,470 cases with
deceased-SSN errors. All estimates in figure 3 have a margin of error of +/-9 percentage points or
fewer at the 95 percent confidence level.
a
In addition to deceased-SSN errors, we also found 24,278 issuances
during fiscal years 2009 and 2010 to individuals who applied for a
passport using a likely invalid SSN that SSA has never issued.
Records that did not fall into any of the two categories above were classified as “other.” Examples
include instances where the applicant did not provide an SSN, or where GAO was unable to ascertain
the applicant’s actual SSN.
34
34
State may issue multiple passports to the same individual, such as when the applicant
applies for both a passport book and a passport card. These 24,278 passport issuances
were associated with 22,543 unique SSNs. In some cases, more than one individual used
the same likely invalid SSN to apply for a passport.
For
example, SSA has never issued the SSN “999-99-9999,” so we would
Page 29 GAO-14-222 Department of State Passport Services
have included an applicant who used that SSN in this population.
35
We
randomly selected a nongeneralizable sample of 15 cases from the
population of unique, likely invalid SSNs for additional review (see fig.
11). In seven cases, State improperly recorded the applicants’ SSNs in
TDIS. In six cases, the applicant provided an incorrect SSN, five of which
were close to the applicant’s actual SSN.
36
Figure 11: Summary of Analysis and Invalid Social Security Number Sample
We did not determine the
cause of the invalid SSNs in the remaining two cases, because they
involved minors for whom we could not ascertain the passport recipient’s
actual SSNs.
a
State officials were unable to identify the specific reason for the
deceased-SSN errors or likely invalid SSNs we identified, but the agency
said it has taken actions to capture correct SSN information more
consistently. For example, since January 2010, State’s management has
issued memorandums clarifying the policies and procedures for capturing
These totals are solely based on the matching criteria described. We conducted additional reviews to
verify data for the sample items in the next column. We did not verify that all issuances from our
matching analysis were actual cases of fraud or issuances to individuals with active warrants. Rather,
we selected samples for additional review, and referred all matches to State for further investigation.
35
26 U.S.C. § 6039E requires passport applicants to provide an SSN, if they have one,
when applying for or renewing a passport. When an applicant has not been issued an
SSN, State requests applicants to enter all zeroes in the SSN field on the passport
application. We therefore excluded SSNs with all zeroes from our analysis of invalid
SSNs. However, we included in our analysis other single-character SSNs, such as “999-
99-9999.” While these SSNs are invalid, they may reflect an attempt by applicants without
SSNs to follow State’s guidance.
36
We defined a “close” SSN as one where at least six digits matched the applicant’s actual
SSN. SSNs that were not close were defined as those with more than three digits that did
not match the applicant’s actual SSN.
Page 30 GAO-14-222 Department of State Passport Services
and correcting applicant SSN information. In addition, in September 2011,
State signed an information-exchange agreement with SSA to implement
a real-time verification process through a secure online system. As of
June 2013, State officials said all 28 domestic agencies and centers were
able to verify SSNs in real time. We have not assessed these measures’
effects on reducing SSN errors, but among the issuances we reviewed,
SSN errors fell from about 24,762 in fiscal year 2009 to approximately
12,986 in fiscal year 2010.
37
We provided a draft of this report to State and the Department of Justice
for comment. State provided technical comments, which we incorporated
into the report, as appropriate. The Department of Justice did not have
any comments.
As agreed with your offices, unless you publicly announce the contents of
this report earlier, we plan no further distribution until 30 days from the
report date. At that time, we will send copies to interested congressional
committees, the Secretary of State, and the Attorney General. In addition,
the report is available at no charge on the GAO website at
http://www.gao.gov.
If you or your staff have any questions about this report, please contact
me at (202) 512-6722 or lor[email protected]. Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page
of this report. GAO staff who made major contributions to this report are
listed in appendix V.
Stephen M. Lord
Managing Director
Forensic Audits and Investigative Service
37
These incorrect SSNs only include SSNs associated with a deceased individual, or
SSNs that we could identify as having never been issued. They do not include legitimately
issued SSNs listed incorrectly in State’s data that do not belong to a deceased individual,
and therefore the count may be understated.
Agency Comments
Appendix I: Objectives, Scope, and
Methodology
Page 31 GAO-14-222 Department of State Passport Services
You asked that we assess potential fraud in the Department of State’s
(State) passport program. This report examines potentially fraudulent or
high-risk issuances among passports issued during fiscal years 2009 and
2010.
1
To examine potentially fraudulent and high-risk passport issuances in
fiscal years 2009 and 2010,
2
we matched State’s passport-issuance data
for approximately 28 million passport issuances to databases containing
information about individuals who were (1) deceased, (2) incarcerated in
a state prison facility, (3) in the custody of the federal Bureau of Prisons
(BOP), or (4) the subject of an active warrant at the time of the passport
issuance.
3
1
For purposes of this report, potentially fraudulent passport issuances are those that
involve an applicant using someone else’s identity to apply for and receive a passport. We
defined high-risk passport issuances as issuances to individuals who may pose a risk to
public safety, but who did not necessarily steal someone’s identity to apply for a passport,
such as people with active warrants for felony charges.
We conducted this matching on the basis of common data
elements including Social Security number (SSN), name, and date of
birth. We also analyzed the passport data to identify issuances to
applicants who provided an invalid SSN, which was defined as an SSN
that had not been assigned at the time of the passport application, or had
a high risk of misuse. Because our review focused on passport issuances
with certain fraud indicators, we did not review other types of passport
fraud, such as identity theft involving individuals who were not deceased
or imprisoned. Similarly, since we focused on high-risk issuances
involving applicants with active warrants, we did not match the passport
data against all databases with individuals at risk of misusing a passport,
such as the Federal Bureau of Investigation’s Terrorist Screening Center
data. We only reviewed data with passport issuances, therefore we did
not examine passport applications that were rejected or abandoned by
the applicant. In addition, our unit of analysis was passport issuances,
2
We used data from these fiscal years because they were the most recent, full fiscal years
available at the time State complied with our data request.
3
According to publically available passport issuance statistics, State issued a combined
total of 28,964,775 passports during fiscal years 2009 and 2010. GAO reviewed
domestically issued passports and excluded passports issued by the Special Issuance
Agency to government travelers. We reviewed a total of 28,000,063 passport issuance
records. State erroneously excluded passport records from the Houston Passport Agency
for the months of October 2008, November 2008, and December 2009 and provided
duplicate passport records from the Houston Passport Agency for the months of
November 2009 and September 2010.
Appendix I: Objectives, Scope, and
Methodology
Appendix I: Objectives, Scope, and
Methodology
Page 32 GAO-14-222 Department of State Passport Services
instead of passport holders. Some individuals in our sample may have
been issued multiple passports.
In addition, we examined policies, guidance, including the Foreign Affairs
Manual, and other materials provided to passport specialists. We
reviewed changes to State’s controls since fiscal years 2009 and 2010
with respect to preventing certain fraudulent or high-risk passport
issuances. We also assessed the reliability of State’s passport data,
TECS travel activity data provided by the Financial Crimes Enforcement
Network (FinCEN), the Social Security Administration’s (SSA) full death
file, prisoner databases provided by the BOP and by departments of
corrections in 15 selected states, as well as data on individuals with open
warrants provided by the Marshals Service, by reviewing relevant
documentation, interviewing knowledgeable agency officials, and
examining the data for obvious errors and inconsistencies. We concluded
that all but four of these databases were sufficiently reliable for the
purposes of this report. Through data tests and interviews, we concluded
that state prisoner data from Illinois, Louisiana, Michigan, and
Pennsylvania were not sufficiently reliable for our purposes. We did not
assess the reliability of state prisoner data from North Carolina because
they were not provided in time to be included in our analysis.
To identify individuals using the SSN of a deceased individual, we
matched State passport data to SSA’s full death file as of September
2011. The full death file contains all of SSA’s death records, including
state-reported death information. We included only those individuals who
died more than 120 days before the passport was issued to ensure that
our matches did not include legitimate applicants who died shortly after
submitting an application.
4
From 13,470 passport records with names that did not match the
corresponding record in the full death file, we selected a stratified random
sample that consisted of 140 passport issuances, evenly divided between
fiscal years 2009 and 2010. For each case, we examined a copy of the
original passport application and submitted the SSN from State’s passport
We further divided these passport applications
into two groups on the basis of whether the name in the passport file
matched the name in the corresponding death file record.
4
As previously noted, we selected 120 days after death to allow for approximately 60 days
of passport application processing time and 60 days of lag time in reporting an individual’s
death to SSA for inclusion in the full death file.
Appendix I: Objectives, Scope, and
Methodology
Page 33 GAO-14-222 Department of State Passport Services
data to SSA for verification. We analyzed these cases to determine
whether the applicant provided the correct SSN and State recorded it
incorrectly, or whether the applicant provided the wrong SSN and State
recorded the incorrect SSN in its system. Our estimates had a margin of
error of at most +/-9 percentage points for the entire population, at the 95
percent confidence level.
From 181 passport records with names that matched the corresponding
record in the death file, we randomly selected a nongeneralizable sample
of 15 records for additional analysis. To verify that these individuals were
deceased at the time their identity was used to apply for a passport, we
attempted to obtain a death certificate for each applicant. In some cases,
we were unable to obtain a death certificate because we could not identify
the state in which the individual’s death was recorded or because state
officials could not or would not provide the death certificate to us. The
results of this sample are not generalizable to the entire population of
applicants using the SSN and name of a deceased person.
To identify individuals incarcerated at the time of passport issuance, we
matched State passport data to a database of federal prisoners provided
by the BOP and prisoner databases from Alabama, Arizona, California,
Florida, Georgia, Indiana, Missouri, New York, Ohio, Texas, and Virginia.
Federal prisoner data included individuals incarcerated during fiscal years
2009 and 2010. State prisoner data included individuals incarcerated as
of the date the state provided data to us, which ranged from May to
November 2011. We identified records for which the passport applicant’s
SSN, name, and date of birth matched that of a person who was
incarcerated on the date of passport issuance. State prisoner data from
Florida, New York, and Texas did not contain dates of birth. For these
states, we matched passport data to state prison data by SSN and name
only. From our matches, we randomly selected 15 federal prisoners and
up to 2 prisoners incarcerated in each of the states for additional analysis.
If a state had two or fewer valid matches, we selected all matches from
that state, for a total of 14 cases from eight different states. Three states
did not have any matches. We obtained documentation from BOP and
prison officials from the eight states to confirm that the selected
individuals were incarcerated on the dates of passport application and
issuance. The results of these samples are not generalizable to the entire
population of applicants using the name, SSN, or date of birth of an
incarcerated person. However, the cases offered insights on applicants
who potentially used the identity of prisoners to apply for passports, and
related efforts by State to identify such individuals.
Appendix I: Objectives, Scope, and
Methodology
Page 34 GAO-14-222 Department of State Passport Services
To identify individuals with active warrants at the time they applied for a
passport, we matched State passport data to warrant data provided by
the Marshals Service. We identified records for which the passport
applicant’s SSN, name (or alias), and date of birth matched that of an
individual with an open warrant on the date of passport issuance. From
this population, we randomly selected 15 warrants for crimes other than
misdemeanors and traffic violations for additional analysis. We confirmed
with the Marshals Service that all 15 individuals had warrants related to
felony charges. In addition, we referred to documentation provided by the
Marshals Service that had warrant information for each case. We
compared the warrant dates, the SSNs, names, and dates of birth, if
available, in the warrant data we received from the Marshals Service for
our matching analysis to the hard-copy documentation. For all 15 cases,
the warrant issuance dates, as well as the names and fugitive unique
identifiers, in the hard-copy documentation matched the information in the
warrant data we used to match with the passport database. In addition,
for 12 cases, at least one date of birth and SSN noted in the hard-copy
documentation matched the warrant data used for our matching analysis.
The dates of birth and SSNs in the documentation for 3 of the 15
individuals were redacted, and therefore the match was based on name
and unique identifier only. The results of this sample are not generalizable
to the entire population of applicants using the name (or alias), SSN, and
date of birth of a fugitive, but provided insights about State’s efforts to
identify such individuals during the adjudication process.
Because we matched passport data to databases of deceased
individuals, prisoners, and fugitives using two or more identifiers—SSN,
name, date of birth—we are generally confident in the accuracy of our
results. However, in some cases, our matches may include applicants
who were not deceased, incarcerated, or the subject of an active warrant.
This can occur when a passport applicant has an SSN, name, and date of
birth that are similar to an individual listed in one of the other databases
or when the applicant is listed in the other database erroneously. In
addition, our matches may be understated because we may not have
detected applicants whose identifying information in the passport data
differed slightly from their identifying information in other databases.
Moreover, federal warrant data do not contain information on all
individuals with an open warrant issued by a state court.
We analyzed State passport data to identify issuances to individuals
using an invalid SSN. We defined invalid SSNs as SSNs that had not
been issued as of fiscal year 2010 or commonly misused SSNs. SSNs
with certain digit combinations, such as those starting with 000 or 666,
Appendix I: Objectives, Scope, and
Methodology
Page 35 GAO-14-222 Department of State Passport Services
had never been issued as of fiscal year 2010. Commonly misused SSNs
include single-character SSNs, or SSNs that have been publically
disclosed in advertisements. From this population, we selected 15 cases
for additional analysis. The results of this sample are not generalizable to
the entire population of applicants using an invalid SSN, but the cases
provided insights on State’s controls related to identifying inaccurate
SSNs on passport applications. In total, we selected 214 cases for
additional analysis, as shown in figure 12.
Figure 12: Summary of Matching Analysis and Samples by Population
a
These totals are solely based on the matching criteria described. We conducted additional reviews to
verify data for the sample items in the next column. We did not verify that all issuances from our
matching analysis were actual cases of fraud or issuances to individuals with active warrants. Rather,
we selected samples for additional review, and referred all matches to State for further investigation.
Appendix I: Objectives, Scope, and
Methodology
Page 36 GAO-14-222 Department of State Passport Services
b
In our data matching, we used state prisoner data from 11 states including Alabama, Arizona,
California, Florida, Georgia, Indiana, Missouri, New York, Ohio, Texas, and Virginia. However, even
though we used data from all 11 states during our matching analysis, three of these state prison
databases did not have any valid matches to the passport data. In addition, data from three states did
not include dates of birth, and therefore records in these databases matched to State’s passport data
based on Social Security number (SSN) and name only.
c
As described in more detail later in this report, our review indicated that some matches may be a
result of identity theft perpetrated by the state prisoner, prior to incarceration, and not the passport
applicant.
d
We selected a nongeneralizable sample of up to two prisoners incarcerated in each of the states we
reviewed for a total of 14 cases from eight different states.
e
This number includes passport issuances to people residing in halfway houses, and therefore may
not represent issuances to individuals using the identities of federal prisoners to apply for a passport.
f
These 486 passport issuances were associated with 442 unique individuals that had a total of 564
open warrants. We did not confirm that all of these warrants were associated with felony charges, but
we excluded warrants with a description of either a “traffic crime” or “misdemeanor” from our analysis.
g
In all, we selected a total of 214 passport issuances for additional review
for our five nongeneralizable and one generalizable samples. For each of
the 214 passport issuances selected, we reviewed a copy of the original
passport application, verified the SSN in State’s passport data using
SSA’s database, and obtained records of the passport holder’s travel
activity from FinCEN. We also reviewed State documentation of additional
investigative activities taken in any of our cases. Where applicable, we
obtained additional documentation about the death, incarceration, or
fugitive status of applicants from federal and state agencies, and follow-
up actions planned or taken by State. We could not determine whether
the passport issuance was inappropriate or fraudulent without additional
investigation of the facts and circumstances for each individual case. We
were not able to perform these investigations due to restrictions on the
use of State’s passport data. The period required for our review was a
result of various factors, including a data-sharing negotiation with State,
time required to receive requested data and documentation, extensive
data preparation and analysis involving multiple databases, and the
necessity for resource-intensive reviews of information on-site, given the
sensitivity of certain information.
The generalizable stratified random sample of 140 passport issuances included 70 passport
issuances from both fiscal years 2009 and 2010.
We performed this audit from March 2010 through May 2014 in
accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe that
the evidence obtained provides a reasonable basis for our findings and
conclusions based on our audit objectives.
Appendix II: Passport Application and
Adjudication Process
Page 37 GAO-14-222 Department of State Passport Services
This figure is a printable version of the interactive graphic presented
above, which provides an overview of the passport application and
adjudication process.
Appendix II: Passport Application and
Adjudication Process
Appendix II: Passport Application and
Adjudication Process
Page 38 GAO-14-222 Department of State Passport Services
Figure 13: Passport Application and Adjudication Process
Appendix III: The Department of State’s (State)
Use of the Social Security Administration’s
(SSA) Records for Death Checks
Page 39 GAO-14-222 Department of State Passport Services
In April 2007, State and SSA signed an information-exchange agreement
that allows State to automatically query SSA’s Enumeration Verification
System records for verifying applicants’ identities and identifying
deceased individuals. SSA provides State the death status of an applicant
when all identifying fields—the Social Security number (SSN), name, and
date of birth on the passport application—match an SSA record.
1
1
SSA’s criteria for matching include tolerances on the name and date of birth to account
for input or typographical errors on records submitted.
However, if one of these identifying elements does not match, SSA will
not provide a response with respect to death status. For example, SSA
would not provide State a death status if an applicant submitted the SSN
of a deceased individual but used a different name. Figure 14
summarizes the potential responses State receives from SSA.
Appendix III: The Department of State’s
(State) Use of the Social Security
Administration’s (SSA) Records for Death
Checks
Appendix III: The Department of State’s (State)
Use of the Social Security Administration’s
(SSA) Records for Death Checks
Page 40 GAO-14-222 Department of State Passport Services
Figure 14: Social Security Administration’s Responses and Actions of Passport Specialists
Appendix IV: Description of Warrant-Matching
Analysis
Page 41 GAO-14-222 Department of State Passport Services
Out of the approximately 28 million combined passport issuances we
reviewed from fiscal years 2009 and 2010, we identified 1,096 individuals
whose Social Security number (SSN), name, and date of birth were
associated with 1,309 warrants in data provided by the Marshals Service
from the Justice Detainee Information Center database. Of these 1,309
warrants, we identified 486 passport issuances to individuals who may
have had active felony warrants at the time of application submission.
Over half of the 1,309 warrants were for offenses such as misdemeanors
and traffic violations. According to officials at the Department of State
(State), the department has no regulatory authority to take action on
misdemeanor warrants.
Officials from the Marshals Service said the agency tracks other warrants
for federal agencies, including those related to misdemeanors and traffic
offenses, as well as cases at the state level that require the assistance of
the Marshals Service. Officials of the Marshals Service also said they do
not have responsibility for entering information about such warrants into
the National Crime Information Center (NCIC) database, and therefore
such warrants are classified as non–Class 1 warrants. On the other hand,
Class 1 warrants are those that the Marshals Service is responsible for
entering into NCIC and that it transmits to State’s Consular Lookout and
Support System (CLASS) for adjudication.
Marshals Service officials told us that of the 1,309 warrants we identified
in our matching analysis, 111 (9 percent) were Class 1 warrants. Because
the Marshals Service only transmits Class 1 warrants into CLASS, State
would have identified the non–Class 1 warrants only if it received them
from another source, such as warrants provided by the Federal Bureau of
Investigation. Table 1 lists the types of open warrants we identified from
our matching analysis.
1
1
The warrant data we received from the Marshals Service included warrants classified as
“State/Local” that may have included both felonies and misdemeanors. In addition, the
underlying offense may have been a misdemeanor or traffic violation for some warrants
that were parole violations; we were unable to identify these on the basis of the data
provided.
Appendix IV: Description of Warrant-
Matching Analysis
Appendix IV: Description of Warrant-Matching
Analysis
Page 42 GAO-14-222 Department of State Passport Services
Table 1: Warrants by Type of Offense to Individuals Issued Passports from Fiscal
Years 2009 and 2010
Offense type
Number of active
warrants
a
Percentage
Traffic Crimes and Misdemeanors 766 59%
Others 112
b
9
Parole Violation and Bond Default 110 8
Drug Crimes 101 8
Weapon Offenses and Violent Crimes 79 6
Fraud, Forgery, and Financial Crimes 66 5
Sex Offenses 65 5
Border and Immigration Crimes 10 < 1
Total 1,309 100%
Source: GAO analysis of data provided by the Department of State (State) and the Marshals Service
a
One individual can have more than one warrant. The table includes warrants for 1,096 different
fugitives.
b
Some of the 1,309 passport issuances we identified with active warrants
were associated with felony offenses, including violent crimes ranging
from assault to homicide. Ten of the warrants were associated with
border and immigration crimes such as smuggling aliens and immigration
violations. State may have identified and reviewed these warrants during
the adjudication process, and decided to issue the passport after
reviewing and resolving the circumstances of the case. We referred all
our matches to State for further review and investigation.
This offense type includes various crimes, such as gambling, burglary, and child neglect.
Appendix V: GAO Contact and Staff
Acknowledgments
Page 43 GAO-14-222 Department of State Passport Services
In addition to the contact above, Heather Dunahoo (Assistant Director);
Hiwotte Amare; James Ashley; Patricia Donahue; Richard Hillman; Leslie
Kirsch; Maria McMullen; Sandra Moore; Anthony Moran; James Murphy;
Rebecca Shea; and Gavin Ugale made key contributions to this report.
Appendix V: GAO Contact and Staff
Acknowledgments
GAO Contact
Staff
Acknowledgments
(192352)
The Government Accountability Office, the audit, evaluation, and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding decisions.
GAO’s commitment to good government is reflected in its core values of
accountability, integrity, and reliability.
The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO’s website (http://www.gao.gov). Each weekday
afternoon, GAO posts on its website newly released reports, testimony,
and correspondence. To have GAO e-mail you a list of newly posted
products, go to http://www.gao.gov and select “E-mail Updates.”
The price of each GAO publication reflects GAO’s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO’s website,
http://www.gao.gov/ordering.htm.
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional information.
Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts .
Visit GAO on the web at www.gao.gov.
Contact:
Website: http://www.gao.gov/fraudnet/fraudnet.htm
E-mail: f[email protected]
Automated answering system: (800) 424-5454 or (202) 512-7470
Katherine Siggerud, Managing Director, [email protected], (202) 512-
4400, U.S. Government Accountability Office, 441 G Street NW, Room
7125, Washington, DC 20548
Chuck Young, Managing Director, young[email protected], (202) 512-4800
U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, DC 20548
GAO’s Mission
Obtaining Copies of
GAO Reports and
Testimony
Order by Phone
Connect with GAO
To Report Fraud,
Waste, and Abuse in
Federal Programs
Congressional
Relations
Public Affairs
Please Print on Recycled Paper.
