Military Personnel: DOD and the Services Need to Take Additional 
Steps to Improve Mobilization Data for the Reserve Components	 
(20-SEP-06, GAO-06-1068).					 
                                                                 
GAO has previously reported on the Department of Defense's (DOD) 
ability to track reservists deployed to the theater of operations
and made recommendations. Reliable mobilization and deployment	 
data are critical for making decisions about reserve force	 
availability and medical surveillance. Because of broad 	 
congressional interest, GAO initiated a review under the	 
Comptroller General's authority to conduct evaluations on his own
initiative to determine (1) what DOD data indicate are the number
of reservists mobilized and deployed in support of the Global War
on Terrorism (GWOT) and the selected demographic and deployment  
characteristics of those deployed and (2) whether DOD's reserve  
deployment and mobilization data and analyses are reliable. GAO  
analyzed data and data analyses from DOD's Contingency Tracking  
System (CTS) and interviewed agency officials.			 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-06-1068					        
    ACCNO:   A61346						        
  TITLE:     Military Personnel: DOD and the Services Need to Take    
Additional Steps to Improve Mobilization Data for the Reserve	 
Components							 
     DATE:   09/20/2006 
  SUBJECT:   Armed forces reserves				 
	     Data collection					 
	     Data integrity					 
	     Defense capabilities				 
	     Internal controls					 
	     Military demographics				 
	     Military forces					 
	     Military personnel 				 
	     Military personnel deployment			 
	     Military reserve personnel 			 
	     Mobilization					 
	     Defense Manpower Data Center			 
	     Global War on Terrorism				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-06-1068

     

     * Results in Brief
     * Background
     * DOD Data Show Demographic and Deployment Characteristics of
          * DOD Data on Selected Demographic and Deployment Variables
               * Number of Reservists Mobilized and Deployed
               * Number of Reservists Who Have Deployed One, Two, or Three or
               * Number of Reservists Who Have Spent 1 Year or Less, 1 to 2 Y
               * Reserve Component Categories for Reservists Deployed in Supp
               * Citizenship Status of Reservists at the Time of Their Most C
               * Race and Ethnicity for Reservists Deployed in Support of GWO
               * Gender of Reservists Deployed in Support of GWOT
               * State of Residence of Reservists Deployed in Support of GWOT
               * Occupational Information about Enlisted and Officer Reservis
          * Data for the Volunteer Status, Location Deployed, and Unit D
     * DOD and the Services Have Taken Steps to Improve the Reliabi
          * DMDC and the Services Are Updating the Mobilization Data in
          * DOD Has Not Fully Addressed Data Issues That Could Improve D
          * DMDC Does Not Have Effective Controls for Ensuring the Accur
     * Conclusions
     * Recommendations for Executive Action
     * Agency Comments and Our Evaluation
     * Appendix I: Scope and Methodology
     * Appendix II: Data on Reservists Deployed in Support of GWOT
     * Appendix III: Comments from the Department of Defense
     * Appendix IV: GAO Contact and Staff Acknowledgments
          * GAO Contact
          * Acknowledgments
          * Obtaining Copies of GAO Reports and Testimony
               * Order by Mail or Phone
          * To Report Fraud, Waste, and Abuse in Federal Programs
          * Congressional Relations
          * Public Affairs

Report to Congressional Committees

United States Government Accountability Office

GAO

September 2006

MILITARY PERSONNEL

DOD and the Services Need to Take Additional Steps to Improve Mobilization
Data for the Reserve Components

GAO-06-1068

Contents

Letter 1

Results in Brief 3
Background 6
DOD Data Show Demographic and Deployment Characteristics of Hundreds of
Thousands of Reservists Deployed in Support of GWOT 10
DOD and the Services Have Taken Steps to Improve the Reliability of
Reserve Component Data, but More Needs to Be Done 25
Conclusions 34
Recommendations for Executive Action 35
Agency Comments and Our Evaluation 35
Appendix I Scope and Methodology 42
Appendix II Data on Reservists Deployed in Support of GWOT through June
30, 2006 47
Appendix III Comments from the Department of Defense 51
Appendix IV GAO Contact and Staff Acknowledgments 54

Tables

Table 1: Citizenship Status of Reservists Deployed in Support of the
Global War on Terrorism by Fiscal Year through June 30, 2006 20
Table 2: Race of Reservists Deployed in Support of the Global War on
Terrorism through June 30, 2006 21
Table 3: Ethnicity of Reservists Deployed in Support of the Global War on
Terrorism through June 30, 2006 21
Table 4: Gender of Reservists Deployed in Support of the Global War on
Terrorism by Reserve Component through June 30, 2006 22
Table 5: Reserve Component Categories for Reservists Deployed in Support
of the Global War on Terrorism by Fiscal Year through June 30, 2006 47
Table 6: State, Territories, and Armed Forces Areas of Residence for
Reservists Deployed in Support of the Global War on Terrorism through June
30, 2006 48

Figures

Figure 1: Organizational Hierarchy for DOD's Defense Manpower Data Center
9
Figure 2: Reservists Mobilized and Deployed in Support of the Global War
on Terrorism through June 30, 2006 12
Figure 3: Reservists Mobilized in Support of the Global War on Terrorism
by Fiscal Year through June 30, 2006 13
Figure 4: Reservists Deployed in Support of the Global War on Terrorism by
Fiscal Year through June 30, 2006 14
Figure 5: Number of Reservists Who Have Deployed One, Two, or Three or
More Times in Support of the Global War on Terrorism through June 30, 2006
16
Figure 6: Number of Reservists Who Have Spent 1 Year or Less, 1 to 2
Years, or More Than 2 Years Deployed in Support of the Global War on
Terrorism through June 30, 2006 17
Figure 7: Percentage of Reservists Deployed in the Selected Reserve in
Support of the Global War on Terrorism through June 30, 2006 18
Figure 8: Citizenship Status of Reservists at the Time of Their Most
Current Deployment in Support of the Global War on Terrorism through June
30, 2006 19
Figure 9: Number of Reservists Deployed in Support of the Global War on
Terrorism by State through June 30, 2006 23

Abbreviations

CTS Contingency Tracking System

DEERS Defense Eligibility Enrollment Reporting System

DFAS Defense Finance and Accounting Service

DIMHRS Defense Integrated Military Human Resource System

DMDC Defense Manpower Data Center

DOD Department of Defense

GWOT Global War on Terrorism

RCCPDS Reserve Components Common Personnel Data System

USD (P&R) Under Secretary of Defense, Personnel and Readiness

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

United States Government Accountability Office

Washington, DC 20548

September 20, 2006

Congressional Committees

Since President Bush signed an Executive Order establishing the Global War
on Terrorism (GWOT)1 on September 14, 2001, hundreds of thousands of
National Guard and Reserve servicemembers2 have been activated, mobilized,
and deployed3 in support of efforts in, among other places, Afghanistan
and Iraq. As the Department of Defense (DOD) continues to rely on about
1.3 million reservists to carry out its military operations domestically
and abroad, there continues to be congressional interest in the impact of
GWOT on reserve employment, income change, medical and health status of
reservists, and other issues.

For decades, DOD has been collecting and reporting information on active
duty and reserve component servicemembers. However, it was not until
October 2001, less than a month after the terrorist attacks, that DOD
emphasized the need for the services to specifically report information
about mobilized and deployed reservists who support contingencies.4 While
DOD has been collecting this information, several reports have emphasized
information about reservists who have been mobilized, not deployed.
Further, some of our prior reports5 have raised concerns about DOD's
ability to effectively track reservists who are being deployed to the
theater of operation. Information about reservists' deployments is needed
to assess reserve force availability and to link reservists' locations
with exposure to medical hazards. Our past work has also confirmed that it
is critical that DOD collect, maintain, and report reliable information on
deployed reservists. In our experience, the data that DOD has reported in
the past about the number of reservists who have been mobilized and
deployed have not been consistent because, for example, the data used came
from different or varied sources and the analyses performed were based on
different analytical assumptions.

1 GWOT includes missions such as Operation Enduring Freedom with
operations in and around Afghanistan and Operation Iraqi Freedom with
operations in and around Iraq.

2 National Guard and Reserve servicemembers include the collective forces
of the Army National Guard and the Air National Guard, as well as the
forces from the Army Reserve, the Navy Reserve, the Marine Corps Reserve,
and the Air Force Reserve. They are known collectively as the reserve
component. This report does not address the Coast Guard Reserve because it
comes under the day-to-day control of the Department of Homeland Security
rather than the Department of Defense (DOD). The Coast Guard does,
however, assist DOD in meeting its commitments and DOD data indicate that
7,053 Coast Guard reservists were mobilized and 204 were deployed in
support of GWOT through June 30, 2006.

3 According to the Department of Defense Dictionary of Military and
Associated Terms, a servicemember is activated when he or she is ordered
to full-time duty in the active military of the United States. A
servicemember is mobilized when he or she becomes part of the process of
assembling and organizing personnel and equipment, activating or
federalizing units and members of the National Guard and Reserves for
active duty, and bringing the armed forces to a state of readiness for war
or other national emergency. A servicemember is deployed when he or she
becomes part of the process to relocate forces and materiel to desired
operational areas. For the purposes of this report, the term mobilized
will refer to reservists who have been activated, mobilized, or both in
support of GWOT.

4 Department of Defense, the Under Secretary of Defense, Personnel and
Readiness, memorandum, "Reporting of Personnel Data in Support of the
National Emergency by Reason of Certain Terrorist Attacks," October 4,
2001.

This report, initiated under the Comptroller General's authority to
conduct evaluations on his own initiative, addressed the following issues:
(1) what DOD data indicate are the number of reservists mobilized and
deployed in support of GWOT and the selected demographic and deployment
characteristics of those deployed and (2) whether DOD's reserve deployment
and mobilization data and analyses are reliable.

For this report, we used data provided by the Defense Manpower Data Center
(DMDC), which is DOD's repository for departmentwide data. We outline the
major assumptions we used to analyze the data in the scope and methodology
section of this report (see app. I). Specifically, to address our
objectives, we obtained and analyzed data from DMDC's Contingency Tracking
System (CTS). CTS is DMDC's system that brings together data about GWOT
from many sources and, according to a senior DMDC official, is the only
source of these data within DOD. The Joint Staff's Manpower and Personnel
office is working toward using only CTS data to determine reserve force
availability for future operations. We also performed reliability
assessments on the data after obtaining an understanding of the data file
structure and the sources of the data. Specifically, we (1) performed
electronic testing of the data files for completeness (that is, missing
data), out-of-range values, and dates outside of valid time frames; (2)
assessed the relationships among data elements; and (3) worked with agency
officials to identify data problems, such as which variables may be
unreliable. We also analyzed the extent to which data provided by DMDC
changed during this review as a result of DMDC's data cleaning effort,
known as a rebaselining. Using applicable guidance,6 we interviewed
knowledgeable officials at DMDC about internal control procedures and
other matters. In addition to the officials at DMDC, we also interviewed
knowledgeable officials from the services, the Office of the Assistant
Secretary of Defense for Reserve Affairs, and the Joint Chiefs of Staff
Manpower and Personnel office. The data we report are sufficiently
reliable for our purposes (that is, providing descriptive information)
with one caveat. Since the Army is in the process of updating its
mobilization data, we could not assess the reliability of the Army's
mobilization data to the same extent as those of the other services. In
comparing our analyses of the data with the analyses reported by DMDC, we
determined that DMDC's analyses were not sufficiently reliable for this
report. We performed our audit work from December 2005 through August 2006
in accordance with generally accepted government auditing standards. A
more thorough description of our scope and methodology is provided in
appendix I.

5 See, for example, GAO, Defense Health Care: Medical Surveillance
Improved Since Gulf War, but Mixed Results in Bosnia, GAO/NSIAD-97-136
(Washington, D.C.: May 13, 1997); Military Personnel: DOD Actions Needed
to Improve the Efficiency of Mobilizations for Reserve Forces, GAO-03-921
(Washington, D.C.: Aug. 21, 2003); Military Personnel: DOD Needs to
Address Long-term Reserve Force Availability and Related Mobilization and
Demobilization Issues, GAO-04-1031 (Washington, D.C.: Sept. 15, 2004); and
Defense Health Care: Improvements Needed in Occupational and Environmental
Health Surveillance during Deployments to Address Immediate and Long-term
Health Issues, GAO-05-632 (Washington, D.C.: July 14, 2005).

                                Results in Brief

Our analysis of DOD data indicates that more than 531,000 reservists have
been mobilized in support of GWOT and more than 378,000 reservists, or
almost 71 percent of the number mobilized, have been deployed in support
of GWOT as of June 30, 2006. The Army National Guard has mobilized and
deployed the greatest number of reservists-more than 230,000 have been
mobilized and more than 163,000 have been deployed. The Navy Reserve had
the fewest number of reservists mobilized-with about 29,000
reservists-while the Marine Corps Reserve had the fewest number deployed
with about 19,000 reservists. The number of reservists mobilized increased
through fiscal year 2003 and has declined since then, while the number of
reservists deployed increased through fiscal year 2003 and then remained
stable through fiscal year 2005. In addition, the majority of reservists
have been deployed once, and of those deployed in support of GWOT,
most-about 307,000 reservists, or 81 percent-have spent a year or less
deployed. Alternatively, more than 65,000 reservists, or 17 percent, have
spent more than 1 year but less than 2 years deployed, and about 6,000
reservists, or fewer than 2 percent, have spent more than 2 years
deployed. The majority of those deployed were part of the Selected
Reserve7 and almost 98 percent were U.S. citizens. Since GWOT began, about
78 percent of reservists who were deployed were White; about 14 percent
were Black or African American; about 2 percent were Asian, Native
Hawaiian, or Other Pacific Islander; and about 1 percent were American
Indian or Alaskan Native. Almost 90 percent identified themselves as
non-Hispanic and 8 percent as Hispanic. Of those deployed, 89 percent were
male and 11 percent were female, and 39 percent came from states in the
southern8 United States, 23 percent from the midwest,9 18 percent from
states in the western10 United States, and 15 percent came from states in
the northeast.11 There were three variables-volunteer status,12 location
deployed, and deploying unit-required by DOD policy for which DMDC could
not provide data either because the data did not exist or because they
were not reliable enough for the purposes of this report.

6 GAO, Standards for Internal Control in the Federal Government,
GAO/AIMD-00-21.3.1 (Washington, D.C.: November 1999).

We found the deployment and mobilization data we used to be reliable for
providing descriptive information. However, the mobilization data, some
deployment data fields, and DMDC's processes for data analyses need
improvement. DMDC and the services have recently taken steps to improve
the reliability of mobilization data; however, additional steps are needed
to make mobilization data and DMDC's analyses of mobilization and
deployment data more reliable, as is required by DOD policy and federal
government internal control standards. DMDC and the services have
undertaken an effort to "rebaseline" or replace all previous
service-provided mobilization data in DMDC's database with new data from
the services. To date, the Navy has validated its reserve component data
file and the Air Force has certified that it has rebaselined its reserve
component data. However, the Army, which has mobilized and deployed the
largest number of reservists, has not completed its rebaselining effort
and has not provided a time frame for doing so. We recognize that the
rebaselining effort is a considerable undertaking replete with numerous
challenges and that it is a positive step in improving the reliability of
the data. However, even if the rebaselining effort were complete,
outstanding issues with certain data definitions across the services would
continue to affect the accuracy and completeness of the data. For example,
the use of the terms "activated," "mobilized," and "deployed" is not
standardized within and among the services. Similarly, there is
variability across the services in the completeness of other variables and
data fields, such as volunteer status, deployment location, and deploying
unit. Along with the rebaselining effort, ensuring that the services
address these data issues and provide all required data to DMDC is an
important step in improving the reliability of the data. With respect to
DMDC's quantitative analyses of its CTS data, DMDC has not documented many
of its procedures, including those for verifying the data analyses it
provides to its customers. Because the data analyses DMDC provided to us
had numerous errors-including overcounting the number of reservists'
deployments as well as overcounting the number of days some reservists
were deployed-and were thus unreliable, we question the effectiveness of
its verification procedures and other supporting internal control
procedures for ensuring accurate reporting. Federal internal control
standards require that data control activities, such as edit checks,
verifications, and reconciliations, be conducted and documented to help
provide reasonable assurance that agency objectives are being met. Until
DOD addresses continuing data definition issues and DMDC documents the
internal control procedures it uses to analyze data and verify its
analyses of data, the information provided to decision makers within
Congress and DOD may be unreliable. Without reliable data and analyses,
decision makers will not be in the best position to make informed
decisions that are grounded in accurate and complete information about
reserve component force availability and medical surveillance issues.

7 The Selected Reserve is composed of those units and individuals
designated by their respective services and approved by the Chairman of
the Joint Chiefs of Staff as so essential to initial wartime missions that
they have priority for training, equipment, and personnel over all other
reserve elements (10 U.S.C. S: 10143).

8 The Census Bureau includes the following states in its definition of the
southern United States: Alabama, Arkansas, Delaware, Florida, Georgia,
Kentucky, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma,
South Carolina, Tennessee, Texas, Virginia, and West Virginia. The
District of Columbia is also included.

9 The Census Bureau includes the following states in its definition of the
midwestern United States: Illinois, Indiana, Iowa, Kansas, Michigan,
Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and
Wisconsin.

10 The Census Bureau includes the following states in its definition of
the western United States: Alaska, Arizona, California, Colorado, Hawaii,
Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.

11 The Census Bureau includes the following states in its definition of
the northeastern United States: Connecticut, Maine, Massachusetts, New
Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont.

12 The legal authority used to involuntarily activate reservists in
response to the terrorist attacks of September 11, 2001, limits the number
that may be involuntarily activated to 1,000,000 (10 U.S.C. S: 12302(c)).

We are making recommendations to the Secretary of Defense to provide
guidance to the services to (1) better define and standardize the use of
key terms, like deployment, and (2) provide all required data, such as
volunteer status and location deployed, to DMDC as well as address data
inconsistencies identified by DMDC. We are also recommending that DMDC
document its internal procedures and processes, including the assumptions
it uses in its data analyses. The Under Secretary of Defense, Personnel
and Readiness provided written comments on a draft of the report. In its
comments, DOD generally concurred with our recommendations, except for the
recommendation to collaborate with other DOD offices on the reasonableness
of the assumptions established and used by DMDC in its data analyses. DOD
stated that DMDC is a support organization and that each organization that
requests reports provides the assumptions that DMDC uses to develop the
reports. However, our audit work showed that DMDC has established and uses
some basic assumptions in analyzing data and that DMDC may not always
discuss these assumptions with other DOD offices, such as Reserve Affairs.
As a result, we continue to emphasize the need for DMDC to document these
assumptions and to collaborate with these offices to ensure a common
understanding of these assumptions. DOD stated it has taken some action on
the other recommendations. DOD also provided technical comments, which we
have incorporated in the report, as appropriate. DOD's comments are
reprinted in their entirety in appendix III of this report.

                                   Background

In 1975, DOD implemented the Reserve Components Common Personnel Data
System (RCCPDS)13 to collect information on current and past members of
the six reserve components-Army National Guard, Air National Guard, Army
Reserve, Navy Reserve, Marine Corps Reserve, and Air Force Reserve. This
information included data on reservists' personal characteristics, such as
name, Social Security number, date of birth, gender, home address, and
education, as well as data on their military characteristics, such as
service, reserve component, prior service status, and date of initial
entry into the reserve forces. According to the director of DMDC, the
services send daily, weekly, and monthly updated data submissions to DMDC
in accordance with applicable guidance.14

13 RCCPDS is an automated information system and associated database that
was established as the official source of statistical tabulation of
reserve component strengths and related data for various users, to include
DOD and Congress.

After the first Gulf War, in a May 15, 1991, memorandum,15 DOD identified
16 recommendations requiring action by many offices within DOD regarding
Desert Storm personnel data issues. For example, the memorandum said that
DOD should consistently report on who participated in the operations and
cites examples of key terms, such as in theater, that were being
interpreted differently by DMDC, the Office of the Secretary of Defense,
and the services. In December 1991,16 DOD reported on how DMDC provided
information about operations Desert Shield and Desert Storm. This report
cited areas for improvement. For example, the report indicated that DMDC
created makeshift procedures to establish and maintain the new data
sources and to accommodate varied data requests. The report cited that
these procedures sometimes resulted in inconsistent or incomplete data
being provided in response to a request. On May 2, 2001, DOD updated
guidance to the military services, among others, to maintain a centralized
database of active duty personnel.17 In this guidance, DOD requires the
services to report personnel information about all active duty military
servicemembers as well as reservists who are ordered to active duty. While
this instruction called for the services to report information about
servicemembers on active duty in support of a contingency, the
requirements for reporting contingency data were not specific.

14 DOD Instruction 1336.5, Automated Extract of Active Duty Military
Personnel Records, May 2, 2001, and DOD Instruction 7730.54, which has
been updated several times since it was first released in 1975.

15 DOD Manpower Data Center Memorandum, "Recommendations Based on Desert
Storm Personnel Data Base Conference of 23-25 April 1991," May 15, 1991.

16 "Defense Manpower Data Center Support for Desert Shield/Desert Storm,"
December 1991.

17 DOD Instruction 1336.5.

On October 4, 2001, the Under Secretary of Defense, Personnel and
Readiness (USD (P&R)),18 issued a memorandum19 that required the services
to report personnel information to DMDC on all active and reserve
component personnel mobilized or deployed in support of GWOT, in
accordance with DOD guidance.20 The purpose of GWOT data was, among other
things, to establish eligibility for benefits and entitlements as a result
of participation in the named contingencies.21 The information is critical
because it provides a historical database with which to assess the impact
of policies and processes, events, and exposures on the health of deployed
reserve component servicemembers. DMDC was tasked with providing reporting
guidance to the services for these data submissions. DMDC sent this
guidance to the services on October 12, 2001.

DMDC is a civilian-led agency with a mission to deliver timely and quality
support to its customers, and to ensure that data received from different
sources are consistent, accurate, and appropriate when used to respond to
inquiries. DMDC reports to the Deputy Under Secretary of Defense for
Program Integration, who is in the Office of the USD (P&R) (see fig. 1).

18 The Under Secretary of Defense for Personnel and Readiness is the
principal staff assistant and advisor to the Secretary of Defense and the
Deputy Secretary of Defense, Total Force Management, as relates to
readiness, National Guard and Reserve component affairs, health affairs,
and personnel requirements and management.

19 Department of Defense, the Under Secretary of Defense, Personnel and
Readiness, memorandum, "Reporting of Personnel Data in Support of the
National Emergency by Reason of Certain Terrorist Attacks," October 4,
2001.

20 DOD Instruction 1336.5.

21 The data are also used for research, actuarial analyses, interagency,
mobilization and contingency reporting, and evaluation of DOD programs and
policies.

Figure 1: Organizational Hierarchy for DOD's Defense Manpower Data Center

In February 2002, USD (P&R) reminded the services in another memorandum of
its earlier requirement for reporting personnel data to DMDC and informed
the services that they had 2 weeks to provide plans to DMDC on how they
were going to correct any personnel data reporting problems. On August 6,
2004, DOD updated prior guidance regarding RCCPDS22 to include an
enclosure23 that set out specific requirements for the services to report
personnel information for all reserve component servicemembers supporting
a named contingency, unlike previous guidance.24 The purpose of the new
enclosure was to ensure more accurate reporting on a named contingency,
such as GWOT missions, as well as to establish eligibility for benefits
and entitlements, and to develop a registry of participants for tracking
in support of research and evaluation of DOD programs and policies.
According to DOD officials, the services, in general, were still reporting
data according to previous guidance for a few years after the new guidance
was issued.

22 DOD Instruction 7730.54, March 15, 1999, was updated with enc. 11 on
August 6, 2004.

23 DOD Instruction 7730.54, enc. 11, August 6, 2004.

24 DOD Instruction 1336.5, enc. 5, May 2, 2001.

In August 2004, DMDC began operation of its CTS database to address DOD's
reporting requirements, including those in the new enclosure (that is,
enclosure 11). The CTS database is DOD's repository for collecting
activation, mobilization, and deployment data for reservists who have
served and continue to serve in support of GWOT. The CTS database contains
both an activation file, which contains mobilization data, and a
deployment file. Both files are updated monthly by service submissions and
cover GWOT from September 11, 2001, to the present. The purpose of the
activation file is to account for and provide medical and educational
benefits for all reservists called to active duty in support of GWOT
contingencies, and it allows DOD to provide data on the number of
reservists who have been mobilized in support of GWOT. The purpose of the
CTS deployment file is to account for a deployed servicemember's
deployment date and location during each deployment event in support of
deployment health surveillance and DOD guidance.25 The database is also
used to track and report the number of reservists who have been deployed
in support of GWOT since September 11, 2001.

    DOD Data Show Demographic and Deployment Characteristics of Hundreds of
              Thousands of Reservists Deployed in Support of GWOT

Our analysis of DOD data indicates that more than 531,000 reservists have
been mobilized in support of GWOT and more than 378,000 reservists, or
about 71 percent of the number mobilized, have been deployed in support of
GWOT through June 30, 2006 (see fig. 2). The Army National Guard deployed
the greatest number of reservists in support of GWOT from September 2001
through June 30, 2006, and, of those, the majority were deployed once. The
data also indicate that the vast majority of reservists who deployed in
support of GWOT were U.S. citizens, White, and male. Further, the data
indicate that most of the reservists spent 1 year or less deployed.

DOD guidance26 requires the services to report timely, accurate, and
complete activation, mobilization,27 and deployment data. DOD guidance
also requires DMDC to collect and maintain mobilization and deployment
data from the services about the reservists. DOD is required by policy to
report personnel data about reservists, such as service, service
component, reserve component category, race, ethnicity, gender,
citizenship status, occupation, unit, and volunteer status regarding a
current mobilization.28 In addition, DOD is required by policy to capture
deployment information such as the location a reservist is deployed to and
the dates the reservist was deployed to that location.29

25 DOD Instruction 6490.03, Deployment Health, August 11, 2006.

26 DOD Instruction 7730.54 and DOD Instruction 6490.03.

27 DMDC's activation and mobilization data are contained in the CTS
activation file. For the purposes of this report, the term mobilized will
refer to reservists who have been activated, mobilized, or both in support
of GWOT.

DOD Data on Selected Demographic and Deployment Variables

  Number of Reservists Mobilized and Deployed

Our analysis of DOD data indicates that more than 531,000 reservists have
been mobilized in support of GWOT and more than 378,000 reservists, or 71
percent of the number mobilized, have been deployed in support of GWOT
through June 30, 2006 (see fig. 2). The number of mobilizations and
deployments peaked in fiscal year 2003 with about 206,000 reservists
mobilized and about 127,000 reservists deployed (see figs. 3 and 4). Since
fiscal year 2003, the total number of mobilizations has declined, while
the number of deployments remained stable through fiscal year 2005. The
Army National Guard has mobilized and deployed the greatest number of
reservists-more than 230,000 mobilized and more than 163,000 deployed. The
Navy Reserve had the least number of reservists mobilized-with about
29,000-while the Marine Corps Reserve had the fewest number deployed with
about 19,000 reservists (see fig. 2). The percentage of the total
reservists mobilized or deployed varies across the fiscal years (see figs.
3 and 4). For example, looking at the percentage of mobilizations by
component each year, Navy Reserve, Air Force Reserve, and Air National
Guard mobilizations occurred early in GWOT and have generally declined
over time. Conversely, the percentage of Army National Guard and Army
Reserve mobilizations has generally increased over time. The greatest
number of Army National Guard deployments-more than 60,000-occurred in
fiscal year 2005 (see table 5 totals in app. II), while also in fiscal
year 2005, the Army National Guard represented the largest deploying
component, with 52 percent of deployments belonging to it (see fig. 4).

28 DOD Instruction 7730.54.

29 DOD Instruction 6490.03.

Figure 2: Reservists Mobilized and Deployed in Support of the Global War
on Terrorism through June 30, 2006

Note: Army mobilization data may change moderately upon completion of the
Army's rebaselining efforts with DMDC. Reservists were only counted once
to identify the total number of individuals who have been mobilized and
deployed in support of GWOT. Totals may not add to 100 percent because of
rounding.

Figure 3: Reservists Mobilized in Support of the Global War on Terrorism
by Fiscal Year through June 30, 2006

Notes: Army mobilization data may change moderately upon completion of the
Army's rebaselining efforts with DMDC. The sum of the number of reservists
mobilized for each fiscal year will be different from the total number of
reservists mobilized for all of GWOT because a reservist can be counted
more than once (that is, for each fiscal year in which he or she began a
mobilization). Totals may not add to 100 percent because of rounding.

aFiscal year 2001 data are for September 11, 2001, through September 30,
2001.

bFiscal year 2006 data are for October 1, 2005, through June 30, 2006.

Figure 4: Reservists Deployed in Support of the Global War on Terrorism by
Fiscal Year through June 30, 2006

Notes: The sum of the number of reservists deployed for each fiscal year
will be different from the total number of reservists deployed for all of
GWOT because a reservist can be counted more than once (that is, for each
fiscal year in which he or she began a deployment). Less than 1 percent of
reservists deployed in fiscal year 2001 were Marine Corps Reserve
servicemembers. Totals may not add to 100 percent because of rounding.

aFiscal year 2001 data are for September 11, 2001, through September 30,
2001.

bFiscal year 2006 data are for October 1, 2005, through June 30, 2006.

  Number of Reservists Who Have Deployed One, Two, or Three or More Times

Although reservists usually deployed only once, some experienced multiple
deployments (see fig. 5). For example, compared to the other reserve
components, the Air National Guard and the Air Force Reserve had nearly
half of their reservists deploying two and three or more times, but they
tend to have shorter deployment cycles according to the Air Expeditionary
Force cycle. Under this cycle, reservists deploy for about 120 days in a
20-month cycle. However, servicemembers assigned to stressed specialties
deploy for longer periods of time and in greater frequency. At the unit
level, some deployment rules have been modified to increase volunteerism
or to add stability to key missions. The Army National Guard and the
Marine Corps Reserve had the lowest percentage of reservists deploying two
and three or more times, but they tend to have longer deployment cycles.
In general, DOD policy30 stipulates that Army units spend 1 year "boots on
the ground" in theater.31 This policy also states that Marine Corps units
below the regimental or group level deploy for 7 months while regimental
and group headquarters units and above deploy for 12 months. This policy
also states that the Chief of Naval Operations' goal is for servicemembers
to have a 6-month deployment with 12 months in a nondeployed status.

30 Department of Defense, Under Secretary of Defense, Personnel and
Readiness, "Action Memorandum: Force Deployment Rules for Operations Iraqi
Freedom and Enduring Freedom," July 30, 2004.

31 "Boots on the ground" is defined in a DOD action memo issued July 30,
2004, as the window of time from when a unit physically arrives in theater
until the unit physically departs from the theater. In addition to time
spent "boots on the ground," Army servicemembers usually have a period
prior to mobilization in which they train at a mobilization station and a
time following deployment where they demobilize.

Figure 5: Percentage of Reservists Who Have Deployed One, Two, or Three or
More Times in Support of the Global War on Terrorism through June 30, 2006

Notes: For the Marine Corps Reserve, very few reservists-less than 1
percent-served three or more deployments. Totals may not add to 100
percent because of rounding.

  Number of Reservists Who Have Spent 1 Year or Less, 1 to 2 Years, or More Than
  2 Years Deployed in Support of GWOT

Our analysis of DOD data indicates that across the services, the majority
of reservists have been deployed once, and of those deployed in support of
GWOT, most-about 307,000 reservists, or 81 percent-have spent a year or
less deployed. Alternatively, more than 65,000 reservists, or 17 percent,
have spent more than 1 year but less than 2 years deployed, and about
6,000 reservists, or fewer than 2 percent, have spent more than 2 years
deployed. The data also indicate that the Marine Corps Reserve had the
highest percentage of reservists serving more than 2 years. In addition,
the data also indicate that very few-less than 1 percent-of Air National
Guard reservists served more than 2 years (see fig. 6).

Figure 6: Percentage of Reservists Who Have Spent 1 Year or Less, 1 to 2
Years, or More Than 2 Years Deployed in Support of the Global War on
Terrorism through June 30, 2006

Notes: For the Air National Guard, very few reservists-less than 1
percent-served more than 2 years. Totals may not add to 100 percent
because of rounding.

  Reserve Component Categories for Reservists Deployed in Support of GWOT

Our analysis of DOD data indicates that most reservists who have deployed
in support of GWOT through June 30, 2006, were members of the Selected
Reserve (see fig. 7 and table 5 in app. II). The majority of units and
individuals in each reserve component are part of the Selected Reserve.
These units and individuals have been designated as so essential to the
initial wartime mission that they have priority for training, equipment,
and personnel over all categories of reservists. Congress authorizes end
strength for Selected Reserve personnel each year. The authorized end
strength for the Army National Guard has been about 350,000 for the past
several years. For fiscal year 2005, data provided by the services to DMDC
indicate that the Army National Guard deployed more than 60,000 Selected
Reserve servicemembers, which represents the highest number of Selected
Reserve servicemembers deployed in a single fiscal year by a single
reserve component since GWOT began.

Figure 7: Percentage of Reservists Deployed in the Selected Reserve in
Support of the Global War on Terrorism through June 30, 2006

Although the services are authorized a maximum number of selected
reservists, the actual number of reservists will fluctuate when additional
reservists are recruited or others leave the reserve component. In
addition, reservists such as those in the Individual Ready Reserve, are
also available for deployment. In general, reservists are trained to have
specific skills and specialties and may not be suited to deploy for a
specific mission until additional training is provided. In addition, some
reservists may not be available for deployment because they are in
training, on medical leave, or awaiting training.

  Citizenship Status of Reservists at the Time of Their Most Current Deployment
  in Support of GWOT

Our analysis of DOD data indicates that almost 98 percent of reservists
who have deployed in support of GWOT through June 30, 2006, were U.S.
citizens at the time of their most current deployment (see fig. 8). The
data indicate that about 1 percent of reservists were non-U.S. citizens or
non-nationals at the time of their most current deployment. The
citizenship status of more than 1,400 reservists was unknown. DOD data
also indicate that 168 reservists' citizenship status changed. Table 1
shows the citizenship status of reservists by reserve component by fiscal
year.

Figure 8: Citizenship Status of Reservists at the Time of Their Most
Current Deployment in Support of the Global War on Terrorism through June
30, 2006

Note: Totals may not add to 100 percent because of rounding.

aU.S. nationals are non-U.S. citizens who owe permanent allegiance to the
United States, such as persons born in American Samoa or Swains Island.

Table 1: Citizenship Status of Reservists Deployed in Support of the
Global War on Terrorism by Fiscal Year through June 30, 2006

                                       Fiscal year                
                             2001a   2002    2003    2004    2005  2006b 
Army     U.S. citizen        3,622  6,868  40,083  47,477  59,513 23,542 
National                                                                 
Guard                                                                    
         U.S. national           0      0       0       0       0      0 
         Non-U.S.               25     54     288     440     708    273 
         citizen or                                                      
         non-national                                                    
         Unknown                 0      0       1       0       0      0 
Air      U.S. citizen        1,469 18,935  21,967  13,086  12,259  9,497 
National U.S.         53   576 633    381     359             376 
Guard    national                                                 
         Non-U.S.      0     3   0      9      11               9 
         citizen or                                               
         non-national                                             
         Unknown       0     1   5      3       1               0 
Army     U.S. citizen        1,158  3,749  35,401  28,793  24,316 12,948 
Reserve  U.S.          0     0   6      4       4               0 
         national                                                 
         Non-U.S.     17    77 766    588     473             201 
         citizen or                                               
         non-national                                             
         Unknown       6 1,379  72     89      84              31 
Navy     U.S. citizen          235  2,464   5,349   5,192   5,168  5,554 
Reserve  U.S.          0     6  15      9      14               9 
         national                                                 
         Non-U.S.      2    23  84     91      77             131 
         citizen or                                               
         non-national                                             
         Unknown       8    87 207    222     158             149 
Marine   U.S. citizen            8  1,252   9,440   5,697   3,177                 895
Corps    U.S.          0     1  33     14      15               0 
Reserve  national                                                 
         Non-U.S.      0    98 479    148     107              34 
         citizen or                                               
         non-national                                             
         Unknown       0     0   0      0       0               0 
Air      U.S. citizen          508 11,630  11,795   8,927   8,407               6,469
Force    U.S.          0     0   0      0       0               0 
Reserve  national                                                 
         Non-U.S.      0     5   4      6       6               7 
         citizen or                                               
         non-national                                             
         Unknown       0     2   4      3       0               0 
Total                        7,111 47,210 126,632 111,179 114,857              60,125

Source: GAO analysis of DOD data.

Note: Reservists can be counted more than once if they deployed more than
once in different fiscal years.

aFiscal year 2001 data are for September 11, 2001, through September 30,
2001.

bFiscal year 2006 data are for October 1, 2005, through June 30, 2006.

  Race and Ethnicity for Reservists Deployed in Support of GWOT

Our analysis of DOD data indicates that about 78 percent of those deployed
for GWOT were White; about 14 percent were Black or African American;
about 2 percent were Asian, Native Hawaiian, or Other Pacific Islander;
and about 1 percent were American Indian or Alaskan Native (see table 2).
Overall, about 5 percent of the deployed reservists declined to indicate
their race. The Army National Guard, the Air National Guard, and the Air
Force Reserve had the highest percentages of the reservists who identified
themselves as White. Further, about 90 percent of those who responded
identified themselves as non-Hispanic and 8 percent as Hispanic (see table
3).

Table 2: Race of Reservists Deployed in Support of the Global War on
Terrorism through June 30, 2006

                        Army      Air                  Marine     Air         
                    National National    Army    Navy   Corps   Force 
                       Guard    Guard Reserve Reserve Reserve Reserve   Total
White             131,686   44,759  66,609  13,905  14,096  23,432 294,487 
Black or African   21,285    3,384  19,541   2,611   1,543   3,191  51,555 
American                                                           
Asian/Native        2,794      592   3,318     492     292     307   7,795 
Hawaiian or                                                        
other Pacific                                                      
Islander                                                           
American            1,193      471     497     138     130     122   2,551 
Indian/Alaska                                                      
Native                                                             
More than one           0      132       0   1,981      64      85   2,262 
race                                                               
Declined to         6,586    1,836   5,127   1,288   3,163   1,632  19,632 
respond                                                            
Total             163,544   51,174  95,092  20,415  19,288  28,769 378,282 

Source: GAO analysis of DOD data.

Note: Although the1997 governmentwide requirements for the collection and
reporting of information on race and ethnicity were to have been
implemented by January 1, 2003, DOD has not yet fully implemented the
requirements and its internal monthly reports continue to use some of the
former racial and ethnic categories.

Table 3: Ethnicity of Reservists Deployed in Support of the Global War on
Terrorism through June 30, 2006

                    Army      Air                      Marine     Air         
                National National    Army    Navy       Corps   Force 
                   Guard    Guard Reserve Reserve     Reserve Reserve   Total
Non-Hispanic  152,350   48,497  84,498  15,003      11,531  26,970 338,849 
Hispanic       11,193    2,579  10,465   1,851       3,054   1,699  30,841 
Unknown             1       98     129   3,561       4,703     100   8,592 
Total         163,544   51,174  95,092  20,415      19,288  28,769 378,282 

Source: GAO analysis of DOD data.

  Gender of Reservists Deployed in Support of GWOT

Our analysis of DOD data indicates that about 338,000 reservists, or about
89 percent of the number deployed, were male (see table 4). About 11
percent of those deployed in support of GWOT were female. Of the
approximately 163,500 Army National Guard servicemembers who have been
deployed through June 30, 2006, more than 92 percent were male. Almost 98
percent of those deployed in support of GWOT through June 30, 2006, for
the Marine Corps Reserve were male, representing the highest percentage of
males compared with females for all of the reserve components.

Table 4: Gender of Reservists Deployed in Support of the Global War on
Terrorism by Reserve Component through June 30, 2006

               Army      Air                                      Air         
           National National    Army    Navy     Marine Corps   Force 
              Guard    Guard Reserve Reserve          Reserve Reserve   Total
Male     150,633   45,674  79,799  17,897           18,806  25,044 337,853 
Female    12,910    5,500  15,276   2,518              482   3,725  40,411 
Unknown        1        0      17       0                0       0      18 
Total    163,544   51,174  95,092  20,415           19,288  28,769 378,282 

Source: GAO analysis of DOD data.

  State of Residence of Reservists Deployed in Support of GWOT

Our analysis of DOD data indicates that California, Texas, Pennsylvania,
and Florida had the highest numbers of reservists who have deployed in
support of GWOT through June 2006 (see table 6 in app. II for the number
of reservists deployed by state of residence by reserve component by
fiscal year). The 4 states combined had more than 76,000 reservists in
residence at the time of their deployments. Eleven states deployed more
than 10,000 reservists each, accounting for more than 160,000 reservist
deployments. Of those deployed, about 39 percent came from states in the
southern United States, about 23 percent from the midwest, about 18
percent from states in the western United States, and about 15 percent
came from states in the northeast part of the country.32 More than 20,000
reservists indicated California or Texas as their state of residence at
the time they were deployed (see fig. 9). Nineteen states and 5
territories had fewer than 5,000 reservists in residence at the time of
their deployment and 20 states and 1 territory had from 5,000 to 9,999
reservists in residence at the time of their deployment.

32 Totals may not add to 100 percent because of rounding and because we
did not include the percentage for territories or Armed Forces areas.

Figure 9: Number of Reservists Deployed in Support of the Global War on
Terrorism by State through June 30, 2006

  Occupational Information about Enlisted and Officer Reservists Deployed in
  Support of GWOT

Our analysis of DOD data indicates that since GWOT began, the occupational
areas33 of enlisted reservists deployed in support of GWOT have stayed
somewhat consistent across all services. For example, the Army National
Guard, the Air Force Reserve, and the Marine Corps Reserve have deployed
reservists mostly in infantry occupational areas including such groups as
infantry, air crew, and combat engineering. All six reserve components
have deployed electrical and mechanical equipment repairers, such as
automotive, aircraft, and armament and munitions. Three of the six reserve
components-the Army National Guard, the Army Reserve, and the Marine Corps
Reserve-have deployed reservists who are service and supply handlers, such
as law enforcement and motor transport.

Since GWOT began, the occupational areas most deployed for reserve
component officers have varied, but all reserve components primarily
deployed tactical operations officers, to include ground and naval arms,
helicopter pilots, and operations staff subgroups. The Army National
Guard, the Air National Guard, and the Navy Reserve have deployed
engineering and maintenance officers, such as the communications and radar
and aviation maintenance occupational subgroups. The Air National Guard,
the Army Reserve, and the Air Force Reserve have deployed reservists in
the health care officer occupational areas, including physicians and
nurses. The Army Reserve and the Marine Corps Reserve have deployed supply
and procurement occupational areas that include transportation, general
logistics, and supply occupational subgroups. The Air Force Reserve has
also deployed intelligence officers in occupational subgroups such as
general intelligence and counterintelligence.

Data for the Volunteer Status, Location Deployed, and Unit Deployed Variables
Were Either Not Available or Not Reliable

We were unable to analyze the volunteer status variable because the data
do not exist for all of the reserve components. Similarly, we were unable
to analyze the deployment location and deployment unit variables because
we determined, in agreement with DMDC officials, that the data in these
fields were not reliable. This issue is discussed further below.

33 Occupational areas, groups, and subgroups are defined by DOD's
Occupational Database. Examples given represent DOD's areas, occupational
groups or subgroups.

  DOD and the Services Have Taken Steps to Improve the Reliability of Reserve
                   Component Data, but More Needs to Be Done

While we found selected deployment and mobilization data to be
sufficiently reliable for our purposes (that is, providing descriptive
data), some of the data were not reliable enough for us to report, even
for descriptive purposes. DMDC and the services, as required by DOD
policy, have taken steps to improve the reliability of the mobilization
data; however, more action is needed to improve the reliability of CTS
data and DMDC's analyses of those data. For example, (1) the rebaselining
effort resulted in substantial changes being made to the mobilization
data, and the Army-which has mobilized and deployed the largest number of
reservists for GWOT-has not completed this rebaselining effort, which the
Joint Staff tasked DMDC and the services to do in November 2005; (2) we
identified data issues that DOD has not addressed that could further
improve the reliability of the data, such as standardizing the use of key
terms like deployment; and (3) DMDC does not have effective controls for
ensuring the accuracy of its data analyses used to produce reports as
required by federal government internal control standards. Although DMDC
and DOD have undertaken a major data cleaning-or rebaselining-effort to
improve the reliability of mobilization data, the effort does not address
some fundamental data quality issues. While we recognize that such a
large-scale effort, although replete with challenges, is a positive step
toward better quality data, if data reporting requirements and definitions
are not uniform, and if there are no quality reviews of DMDC's analyses,
some data elements and DMDC's analyses of those data may continue to be
unreliable. A senior DMDC official stated that it emphasizes getting data
to customers in a timely manner rather than documenting the internal
control procedures needed to improve the reliability of the data and the
data analyses produced. However, with proper internal controls, DMDC could
potentially achieve both timeliness and accuracy. Without reliable data
and analyses, DOD cannot make sound data-driven decisions about reserve
force availability. Moreover, DOD may not be able to link reservists'
locations with exposure to medical hazards.

DMDC and the Services Are Updating the Mobilization Data in CTS, but Concerns
Remain

We have found the deployment and mobilization data we used to be
sufficiently reliable for our purposes (that is, providing descriptive
data), and DMDC and the services have recently taken steps to improve the
reliability of mobilization data. However, additional steps are needed to
make mobilization data more reliable. As previously noted, DOD guidance34
requires the services to report timely, accurate, and complete activation,
mobilization,35 and deployment data. DMDC officials responsible for
overseeing the CTS database stated that a rebaseline of the deployment
data was not necessary because the deployment data matched the data in the
Defense Finance and Accounting Service's (DFAS) systems36 by more than 98
percent. Although DMDC and the services rebaselining of the mobilization
data in CTS has resulted in improvements, the Army, which has mobilized
the greatest number of reservists for GWOT, has not completed its
rebaselining effort. A senior-level DMDC official responsible for
overseeing the CTS database said that the mobilization data in the CTS
database prior to the rebaselining effort were less than 80 percent
accurate for the Army, the Navy, and the Air Force, but that the Marine
Corps' data were generally considered to be accurate prior to the
rebaselining effort. The official also stated that DMDC expects that the
mobilization data within the CTS database will be 90 percent accurate
because of this rebaselining effort, which was still ongoing through
August 2006. While we recognize that this is a considerable undertaking,
to date, only the Navy and the Air Force have validated or certified their
mobilization data files. Navy officials said that the Navy has validated
its personnel records and established a common baseline of data with DMDC.
Air Force Reserve officials said that their data within CTS are now 99 to
100 percent accurate. The Chief of the Personnel Data Systems Division for
the Air National Guard certified that although file discrepancies are
still being reconciled, the data that were processed by DMDC on June 11,
2006, were the most accurate activation37 data and that data accuracy will
improve with each future file sent to DMDC. The DMDC official said that
the Marine Corps had only partially completed its rebaselining effort and
would not be finished until the Marine Corps provided its August 2006 data
file in September 2006. The Army National Guard and Army Reserve are still
working to rebaseline their mobilization data, and the Army has not
provided a time frame for completing the effort.

34 DOD Instruction 7730.54 and DOD Instruction 6490.03.

However, we still have concerns regarding the reliability of the
mobilization data, because the scope of the rebaselining effort changed
and the data changed substantially as a result of the rebaselining. At the
beginning of our review, DMDC and the services referred to the
rebaselining effort as a "reconciliation," which, according to a DMDC
official and a Reserve Affairs official, would have resulted in all data
(current and past) being reviewed and corrected as needed. We acknowledge
that some degree of change is expected in any data cleaning effort,
especially with large-scale, multisource collection methods such as DMDC's
data collection process. However, our experience has shown that cleaning
efforts that result in a large degree of change would suggest systematic
error. Such error raises concerns about the reliability of both the
original data and the "cleaned" data. If both the source data and the
cleaned data are populated with the same assumptions and information, any
reconciliation of data points should result in relatively small change
that correct simply for random error, such as from keypunch or data source
errors. However, for some variables, the data changed substantially as a
result of DMDC and the services' rebaselining or data cleaning effort. Our
analysis shows that data from the period of September 2001 through
December 2005 have changed by about 4 percent to as much as 20 percent.
For example:

35 DMDC's activation and mobilization data are contained in the CTS
activation file. For the purposes of this report, the term mobilized will
refer to reservists who have been activated, mobilized, or both in support
of GWOT.

36 The DFAS systems contain data on special pays and allowances provided
to servicemembers including combat zone tax exclusions and imminent danger
pay.

37 For this report, Air Force activation data are the same as mobilization
data.

           o  The number of reservists mobilized for GWOT through December
           2005 went from about 478,000 to about 506,000-an increase of more
           than 27,000 reservists or a change of more than 5 percent.
           o  The Army Reserve data sustained the greatest change during this
           time with a more than 19 percent increase in the number of
           reservists mobilized. The number of mobilized Army National Guard
           reservists increased more than 7 percent. According to a senior
           DMDC official, the Army data are expected to continue to change,
           perhaps substantially enough to require the rebaselining of the
           data again in the future.
           o  The number of Air National Guard reservists mobilized decreased
           by more than 13 percent.
           o  The Navy Reserve, the Marine Corps Reserve, and the Air Force
           Reserve data all changed about 5 percent.

           DOD officials stated that the rebaselining effort occurred because
           the Joint Staff tasked DMDC and the services with ensuring that
           the data the Joint Staff's Manpower and Personnel office was using
           in CTS were the same data as the services were using to determine
           reserve force availability. According to a senior-level DMDC
           official responsible for overseeing CTS, the rebaselining effort's
           scope changed because all of the services agreed that starting
           over and replacing all of the data would make more sense than
           trying to correct transactions already in CTS, because the
           services found errors in the CTS files initially used for the
           reconciliation. Service officials said that some of the data
           discrepancies developed because of a DMDC quality check procedure
           that sometimes resulted in DMDC replacing the service-submitted
           data with data from other sources. DMDC officials said that they
           did this because the services were unable to report some of the
           required CTS data. According to DMDC officials, service
           submissions have become more complete over time, resulting in DMDC
           now using the quality check procedures only to check the data
           rather than to populate the CTS database. This DMDC official
           stated that DMDC expected the data to change substantially based
           on the issues identified with service data during the initial
           reconciliation effort and the subsequent rebaselining effort.

           Because the rebaselining effort is not complete and the Army-which
           has mobilized and deployed the largest number of reservists for
           GWOT-has not finished the rebaselining, we do not know how much
           the data will continue to change as DMDC and the services work to
           finish this effort. DOD data on reservists' mobilizations and
           deployments are important because decision makers at DOD and in
           Congress need the data to make sound decisions about personnel
           issues and for planning and budgeting purposes.

           DOD Has Not Fully Addressed Data Issues That Could Improve Data
			  Reliability
			  
			  Prior to the rebaselining effort, some services recognized that
           there were data issues that needed to be addressed and took steps
           to do so, as DOD guidance38 requires the services to report
           accurate and complete mobilization and deployment data. However,
           some data issues that would ensure more accurate, complete, and
           consistent mobilization and deployment data across the services in
           the future have not been fully addressed by DOD. Some examples of
           data issues being addressed include the following:

           o  The Air Force and the Navy were having difficulty tracking
           mobilizations based on reservists' mobilization orders, which has
           resulted in both services independently working to develop and
           implement systems that write reservists' orders.
           o  The Army Reserve recently began to modify its mobilization
           systems, which Army officials expect will improve the collection
           of reservists' mobilization data.
           o  The Air Force identified problems with the way in which the
           Defense Eligibility Enrollment Reporting System (DEERS) processed
           end dates for reservists' mobilizations, which resulted in some
           reservists not receiving appropriate benefits (for example, dental
           benefits). Air Force officials worked with officials from the
           Office of the Secretary of Defense and DMDC to identify and
           address the data processing logic issues.

           Despite these positive steps, service process improvements are not
           all complete, and further, there has been no comprehensive review
           across DOD to identify data issues that if addressed, could result
           in more complete, accurate, and consistent mobilization and
           deployment data across and within the services. Reserve Affairs
           officials in the office of Reserve Systems Integration said that a
           more sustainable fix to the processes of collecting data is needed
           to ensure that data captured in the future are accurate and more
           efficiently collected. We agree and have identified some issues
           that may continue to affect data reliability, such as the
           following:

           o  The use of terms, such as activated, mobilized, and deployed,
           has not been standardized across the services. Although the
           department has defined these terms in the Department of Defense
           Dictionary of Military and Associated Terms, the terms are used
           differently by the individual services. In the Air Force,
           "activation" can refer to the time when a reservist either
           volunteers or is involuntarily mobilized; however, the term
           "mobilized" refers only to someone who is not a volunteer. Even
           within a single service, these words can have different meanings.
           For example, an Army National Guard official who participated in
           the rebaselining effort said that Army National Guard
           servicemembers who backfill active duty servicemembers are not
           considered deployed since they have not left the United States.
           However, according to this official, some staff in the Army
           National Guard use "deployed" to include reservists who are
           mobilized within the United States.
           o  There is no single data entry process that would minimize the
           potential for contradictory data about reservists in multiple
           systems. Currently, data about reservists are entered separately
           into multiple systems.
           o  There is no mechanism for DMDC to ensure that the services are
           addressing the data inconsistencies DMDC identifies during its
           ongoing, monthly validation process, such as Social Security
           numbers that are duplicated in two reserve components.
           o  DOD has taken an ad hoc, episodic approach to identifying data
           reporting requirements and to addressing data issues. DOD has
           periodically issued policies regarding its need to collect and
           report specific data, such as volunteer status and location
           deployed, about active duty servicemembers and reservists. As a
           result of changing requirements, many of these policies have
           addendums that include these additional data requirements, which
           are not immediately supported by the services' existing systems
           that are used to collect the data. Over time, this has led to
           disjointed policies that overlap and that require the services to
           modify their existing systems and processes, which can take months
           to complete.
           o  There are incomplete data submissions across the services.
           Specifically, data for volunteer status was not available in CTS
           for all service components, and the location deployed and
           deploying unit data were not reliable enough for the purposes of
           this report. Only three of the six reserve components-the Air
           National Guard, the Marine Corps Reserve, and the Air Force
           Reserve-provide information on a reservist's volunteer status,
           which neither we nor DMDC report because it is not available for
           all six components. Further, DMDC officials said that they
           consider CTS location data incomplete although the data are
           improving with each fiscal year. DMDC officials said that most
           unit information is based on the unit a reservist is assigned to
           and may not represent the unit the reservist is currently deployed
           with in theater. For this reason, we did not consider these data
           reliable enough to report.

           A DMDC official stated that DMDC does not have the authority to
           direct the services to correct data errors or inconsistencies or
           to address data issues. DMDC does, however, work with the services
           and tries to identify and address data challenges. According to
           some service officials, the department plans to implement a new,
           integrated payroll and personnel system-Defense Integrated
           Military Human Resource System (DIMHRS)-and that the services have
           been diverting resources needed to modify their existing systems
           and relevant processes to support DIMHRS. However, our past work
           has shown that DOD has encountered a number of challenges with
           DIMHRS, which is behind schedule, and the current schedule has it
           available no sooner than April 2008, when the Army is scheduled to
           begin implementing the system.

           In general, service officials said that they are working to
           collect data on volunteer status, location deployed, and deploying
           unit; however, Air Force officials stated that they do collect
           data on location deployed and deploying unit and that these data
           are accurate and are being provided to DMDC. Army Reserve
           officials stated that they currently do not have plans to collect
           data on volunteer status.

           DMDC Does Not Have Effective Controls for Ensuring the Accuracy of
			  Its Data Analyses Used to Produce Reports
			  
			  DMDC has not documented (1) its procedures for verifying that the
           data analyses it performs are correct and (2) the procedures for
           monthly validation of service data or the procedures used to
           perform analyses of data. Either of these issues could, if
           documented as part of DMDC's verification process, address some of
           our concerns about internal controls. DMDC is required by policy39
           to develop and produce reports about mobilization data and respond
           to requests for information about deployed personnel. DOD policy40
           requires DMDC and the reserve components to ensure the accuracy of
           files and the resulting reports. Federal government internal
           control standards41 require that data control activities, such as
           edit checks, verifications, and reconciliations, be conducted and
           documented to help provide reasonable assurance that agency
           objectives are being met.

           DMDC officials said that they have internal verification
           procedures that require supervisors to review all data analyses
           used to generate reports, although these procedures are not
           documented. Specifically, the supervisors are to review (1) the
           statistical programming code used to generate the data analyses to
           ensure that the code includes the customer's data analyses
           parameters (that is, the assumptions used to produce the analyses)
           and (2) the "totals" generated to ensure that these totals match
           the control totals that show the number of reservists currently or
           ever mobilized or deployed in support of GWOT. DMDC officials
           acknowledge the importance of verifying the accuracy of the data
           analyses prior to providing the reports to customers, and they
           stated that they had verified the accuracy of the analyses
           provided to us. However, we found numerous errors in the initial
           and subsequent analyses we received of the GWOT data through May
           2006, causing us to question whether DMDC verified the data
           analyses it provided to us and, if it did, whether the current
           process is adequate. For example, we found that DMDC had done the
           following:

           o  Counted reservists with more than one deployment during GWOT
           also among those who deployed only once during GWOT, which
           resulted in overcounting the number of reservists' deployments.
           o  Used ethnicity responses to identify race despite having told
           us that the internal policy was changed in 2006 and that this was
           no longer an acceptable practice.
           o  Counted reservists whose ethnicity was "unknown" as
           "non-Hispanic" although "unknown" does not necessarily mean
           someone's ethnicity is "non-Hispanic" and there was a category for
           unknowns.
           o  Repeatedly categorized data based on a reservist's first
           deployment (when there was more than one) despite agreeing to
           modify this analytical assumption so that we could present data by
           the reservist's most current deployment.
           o  Reported thousands of reservists as having changed citizenship
           status during GWOT although, in our analyses, we found that only
           168 reservists had changed status.
           o  Analyzed data by reserve component categories (for example,
           Selected Reserve and Individual Ready Reserve) rather than by
           reserve component as we had asked. By analyzing the number of days
           a reservist was deployed by reserve component category, a
           reservist could be counted multiple times within one component if
           he or she changed category. This error affected the way in which
           the total number of days a reservist was deployed was calculated.
           For example, if the same reservist served 350 days as an Army
           National Guard Selected Reserve member and an additional 350 days
           as an Army National Guard Individual Ready Reserve member, he or
           she would be counted as two reservists who were each deployed for
           less than a year. However, our intent was to report that the same
           individual had been deployed for a total of 700 days. In our
           analysis, all of a reservist's days deployed were totaled and
           counted once for each reserve component, regardless of which
           category he or she belonged to when deployed.
           o  Miscoded the end date for the analysis of how many days
           reservists were deployed for GWOT. This resulted in up to an
           additional 90 days of deployment being counted for reservists who
           were still deployed at the time the data were submitted to DMDC.

           In our discussions with DMDC officials, they readily acknowledged
           that errors had been made, although they stated that the analyses
           had undergone supervisory review prior to our receiving them.
           During these discussions, we also discovered that many of these
           errors occurred because DMDC had not used all of our data analyses
           parameters, although these officials had stated that this was one
           of the verification process steps followed. Although we were able
           to work with DMDC officials and identify the analytical
           assumptions they were going to use to complete our analyses,
           without documented analytical procedures, it is unclear to what
           degree the analyses DMDC provides to other users of the data also
           contain errors since many may not similarly verify the analyses
           provided to them by DMDC.

           In addition, DMDC officials have not documented additional
           processes that would further support a verification process, such
           as (1) the ongoing, monthly validation process of service-provided
           data42 and (2) the procedures to perform analyses and generate
           reports, including the assumptions DMDC uses when producing
           periodic and special reports for customers. In the past, according
           to the services, the ongoing, monthly validation process DMDC used
           resulted in two sets of data-one set of service data and one set
           of DMDC data-that may not have been the same. For example, we were
           told by the Air Force that, in some cases, service data were
           replaced with default values because of a business rule that DMDC
           applied to the data and that this change resulted in errors to the
           service-provided data. These inconsistent data caused the Joint
           Staff to request that the services and DMDC reconcile the data. As
           stated above, there were errors in the analyses performed to
           generate the reports DMDC provided to us, including DMDC's not
           using many of the assumptions we agreed to for the analyses. DMDC
           also made errors that contradicted its own undocumented policy.

           A senior DMDC official said DMDC has not documented these
           procedures because the organization emphasizes getting data and
           reports to its customers in a timely manner rather than preparing
           this documentation. This official said that documentation is not a
           top priority because situations change rapidly, and it would be
           hard to keep these documents up-to-date. The official also said
           that the errors made in the analyses provided to us were caused by
           human error and the need to provide data quickly. Further, the
           DMDC official said that while there are standard data requests
           that are generated frequently, GAO's request was an ad hoc
           request, and the procedures for addressing such requests, in
           practice, are not as well defined. While we agree that our
           requests met DMDC's definition of an ad hoc request, we disagree
           that sufficient time was not allowed for DMDC to prepare the
           analyses. For the initial request, we worked with DMDC over the
           course of about 5 business days to define the analytical
           assumptions that would be used during the analysis. DMDC then took
           about 8 business days to complete the analysis and provide it to
           us.

           DOD data analyses are important because decision makers at DOD and
           in Congress need the data to make sound decisions about reserve
           force availability, medical surveillance, and planning and
           budgeting. In the absence of documented procedures and the
           necessary controls to ensure that they are implemented, it is
           difficult for an organization to ensure that it has established a
           robust process that is being consistently applied and that
           accurate results are being achieved. Joint Staff and Reserve
           Affairs officials are emphasizing the need to use one data source
           for most analyses to further reduce the inconsistencies in data
           analyses because service-produced analyses and DMDC-produced
           analyses could differ if both are not using the same set of data
           and assumptions. Otherwise, it is possible that the data analyses
           provided to decision makers at DOD or in Congress will be
           incomplete and inconsistent. If the data analyses are incorrect,
           users could draw erroneous conclusions based on the data, which
           could lead to policies that affect reservists in unanticipated
           ways.

           Conclusions
			  
			  DOD recognizes the need for accurate, complete, and consistent
           data and data analyses, and it has taken some preliminary, ad hoc
           steps to improve its data, including undertaking a considerable
           effort to rebaseline its mobilization data. It has not, however,
           addressed some of the inconsistencies in data and data analyses
           departmentwide, such as when terms are used differently from one
           service to the next. Further, service officials stated that it is
           anticipated that a lot of these problems will be addressed when
           DIMHRS is implemented. However, the schedule for DIMHRS continues
           to slip, so it is unclear when this solution will be available. We
           recognize that the need for accurate, complete, and consistent
           data and data analyses about reservist mobilization and deployment
           is always important, and even more so during higher levels of
           mobilization and deployment, such as is the case now with GWOT.
           This is especially true since, in general, there are restrictions
           on the maximum length of time a reservist can be involuntarily
           activated. Thus, having accurate and complete data on a
           reservist's status is critical for determining availability for
           future deployments. This is especially true of the CTS data since
           the Manpower and Personnel office in the Joint Staff and the
           Office of the Assistant Secretary of Defense for Reserve Affairs
           mostly use the data found in CTS. These data also help DOD and
           Congress to understand the potential impacts of policy decisions
           as they relate to reservists who are eligible for TRICARE Reserve
           Select and educational benefits based on the number of days a
           reservist is deployed. DOD has not provided guidance to the
           services to better define and standardize the use of key terms.
           DOD also has not collected and maintained all essential data nor
           has it established a process for ensuring that data
           inconsistencies are resolved. Further, DOD has not documented key
           procedures and processes for verifying the data analyses it
           provides to its customers, thus compromising its ability to ensure
           the accuracy, completeness, and consistency of these analyses.
           Until decision makers in DOD and Congress have accurate, complete,
           and consistent data and analyses, they will not be in the best
           position to make informed decisions about the myriad of reserve
           deployment matters.

           Recommendations for Executive Action
			  
			  We recommend that the Secretary of Defense take the following four
           actions:

           o  Direct the Under Secretary of Defense, Personnel and Readiness,
           to provide guidance to the services to better define and
           standardize the use of key terms, like activation, mobilization,
           and deployment, to promote the completeness, accuracy, and
           consistency of the data within CTS.
           o  Direct the service secretaries to (1) take the steps necessary
           to provide all required data to DMDC, such as volunteer status and
           location deployed, and (2) have the services address data
           inconsistencies identified by DMDC.
           o  Direct the service secretaries to establish the needed
           protocols to have the services report data consistent with the
           guidance above.
           o  Direct the Under Secretary of Defense, Personnel and Readiness,
           to require DMDC to document its internal procedures and processes,
           including the assumptions it uses in data analyses. In doing this,
           the Under Secretary of Defense, Personnel and Readiness, should
           collaborate on the reasonableness of the assumptions established
           and used by DMDC in its data analyses with the Office of the
           Assistant Secretary of Defense for Reserve Affairs and the Joint
           Staff.

           Agency Comments and Our Evaluation
			  
			  The Under Secretary of Defense, Personnel and Readiness, provided
           written comments on a draft of this report and stated that we
           changed one of our original audit objectives and did not inform
           the department of this change. We disagree. While the scope of our
           audit did change after our initial notification letter of June 17,
           2005, was sent to DOD, we notified the proper officials of this
           change in a December 2, 2005, email to the agency-designated
           liaison within the DOD Inspector General's office. In this email,
           we specifically said that we would be contacting DMDC and that we
           would be focusing on data for reserve component activation,
           mobilization, and deployment for GWOT. In accordance with
           generally accepted government auditing standards (GAGAS),43 GAO
           analysts are expected, as appropriate, to review an agency's
           internal controls as they relate to the scope of the performance
           audit. Specifically, we are required by GAGAS44 to review the
           reliability of the data and the data analyses provided to us. To
           assess the reliability of data and data analyses, we often review
           an agency's internal controls that are put in place to ensure the
           accuracy of the data and analyses. As we discuss in our report, we
           found the data to be sufficiently reliable for our purposes.
           However, over the course of the work, the analyses of the data
           DMDC provided to us continued to have errors. This raised concerns
           about the adequacy of DMDC's internal controls for preparing and
           verifying these analyses, which DMDC stated were not documented.
           In accordance with GAGAS, when reporting on the results of their
           work, auditors are responsible for disclosing all material or
           significant facts known to them which, if not disclosed, could
           mislead knowledgeable users or misrepresent the results.
           Consistent errors in DMDC's analyses led us to include an audit
           objective on the reliability of the data and the data analyses.

           In its written comments, DOD generally concurred with three of our
           recommendations and did not concur with one of our
           recommendations. DOD also provided technical comments, which we
           have incorporated in the report, as appropriate. Regarding our
           recommendation that DOD provide guidance to the services to better
           define and standardize the use of key terms, DOD stated that this
           requirement has already been addressed because these terms are
           defined. We acknowledged in our draft report that these key terms
           are defined in the Department of Defense Dictionary of Military
           and Associated Terms. However, as we state in our report, our
           audit work indicates that the services are not operationalizing
           the use of the terms in a consistent manner. The intent of our
           recommendation is to have DOD standardize the use of the key terms
           across the services.

           DOD generally concurred with our recommendation that the services
           provide all required data to DMDC and address data
           inconsistencies, and stated that the services have been directed
           to provide all necessary data and are working to address data
           inconsistencies. While we agree that the services are working with
           DMDC to address data inconsistencies with regard to the
           rebaselining of mobilization data, we also identified other data
           inconsistencies that DOD has not addressed, such as Social
           Security numbers that are duplicated in more than one reserve
           component. We agree with DOD that some requirements cannot be
           immediately supported by service data systems and modifications to
           them can take time to complete. However, as our report notes, some
           service officials stated that resources are being diverted from
           these efforts to the DIMHRS program, which we reported is behind
           schedule. We continue to observe the need for the services to
           provide all necessary data, to address these data inconsistencies,
           and to establish needed protocols to have the services report data
           consistent with DOD guidance, especially since the data are used
           to determine reserve force availability and for medical
           surveillance.

           DOD also generally concurred with our recommendation that DMDC
           document its internal procedures and processes, including the
           assumptions it uses in data analyses. In its written comments, DOD
           stated that DMDC is in the process of developing documentation on
           its internal procedures and processes and has a draft that
           addresses the processes used from receipt of the data from the
           service components to the final quality control of the
           consolidated file. DOD also stated that DMDC has a draft product
           regarding many of the data analyses procedures used. During this
           engagement, we asked if these procedures and processes were
           documented. As we say in the report, DMDC stated that they were
           undocumented and that documenting them was not a priority.
           Although DOD stated that it is in the process of drafting these
           procedures and processes, we were never provided a draft of these
           documents. DOD also stated that while DMDC attempts to document
           the assumptions made in resulting report titles and footnotes, the
           disclosure of assumptions used in data analyses remain the
           responsibility of the requester of the data analyses. Although we
           agree that the requesters of the data bear responsibility to
           disclose the analytical assumptions used in the data analyses, our
           audit work indicates that there are basic assumptions that DMDC
           establishes and uses that, if documented and discussed with those
           who request data analyses, would allow the users to understand how
           the information can be used, as well as the limitations of the
           data analyses. For example, during a discussion with a Reserve
           Affairs official, who uses the data analyses provided by DMDC to
           provide information to senior DOD officials, we stated that DMDC
           defaults to using a servicemember's first deployment rather than
           the most current deployment when preparing data analyses. This
           official was unaware that DMDC used this assumption and stated
           that the expectation was that DMDC was using the most current
           deployment to generate the analyses. This official planned to
           discuss this issue with DMDC in the future.

           In its written comments, DOD did not concur with what it
           characterized as our fourth recommendation. Specifically, DOD
           separated a single recommendation into two recommendations. In the
           draft report we sent to DOD, the recommendation read: "We
           recommend that the Secretary of Defense direct the Under Secretary
           of Defense, Personnel and Readiness, to require DMDC to document
           its internal procedures and processes, including the assumptions
           it uses in data analyses. In doing this, the Under Secretary of
           Defense, Personnel and Readiness, should collaborate on the
           reasonableness of the assumptions used by DMDC in its data
           analyses with the Office of the Assistant Secretary of Defense for
           Reserve Affairs and the Joint Staff." DOD stated that DMDC is a
           support organization that generates reports for a multitude of
           organizations and that each organization that requests reports
           provides the assumptions that DMDC uses to develop the reports.
           However, our audit work showed that DMDC has established and uses
           some basic assumptions in analyzing data and that DMDC may not
           always discuss these assumptions with other DOD offices, such as
           Reserve Affairs. As a result, we continue to emphasize the need
           for DMDC to document these assumptions and to collaborate with
           these offices to ensure a common understanding of these
           assumptions. Although DOD organizations can request data analyses
           using multiple assumptions, without written documentation other
           organizations may not be fully aware of the analytical assumptions
           used by DMDC and this may lead to miscommunication and,
           ultimately, the data analyses may not be valid in that it does not
           report what the user intended. We continue to believe that the
           assumptions used need to be documented and discussed with other
           DOD offices as we recommended. Based on DOD's comments, we
           modified this recommendation to clarify our intent.

           We are sending copies of this report to the Secretary of Defense;
           the Secretaries of the Army, the Navy, and the Air Force; the
           Commandant of the Marine Corps; the Under Secretary of Defense,
           Personnel and Readiness; and other interested parties. We will
           also make copies available to others upon request. In addition,
           the report will be available at no charge on the GAO Web site at
           http://www.gao.gov.

           If you or your staff have any questions on this report, please
           contact me at (202) 512-5559 or stewartd@gao. 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 IV.

           Derek B. Stewart Director, Defense Capabilities and Management

           List of Congressional Committees

           The Honorable John Warner Chairman The Honorable Carl Levin
           Ranking Minority Member Committee on Armed Services United States
           Senate

           The Honorable Lindsey Graham Chairman The Honorable Ben Nelson
           Ranking Minority Member Subcommittee on Personnel Committee on
           Armed Services United States Senate

           The Honorable Ted Stevens Chairman The Honorable Daniel K. Inouye
           Ranking Minority Member Subcommittee on Defense Committee on
           Appropriations United States Senate

           The Honorable Duncan L. Hunter Chairman The Honorable Ike Skelton
           Ranking Minority Member Committee on Armed Services House of
           Representatives

           The Honorable John M. McHugh Chairman The Honorable Vic Snyder
           Ranking Minority Member Subcommittee on Military Personnel
           Committee on Armed Services House of Representatives

           The Honorable C. W. Bill Young Chairman The Honorable John P.
           Murtha Ranking Minority Member Subcommittee on Defense Committee
           on Appropriations House of Representatives

           Appendix I: Scope and Methodology
			  
			  Our objectives were to determine (1) what Department of Defense
           (DOD) data indicate are the number of reservists mobilized and
           deployed in support of the Global War on Terrorism (GWOT), and the
           selected demographic and deployment characteristics of those
           deployed and (2) whether DOD's reserve deployment and mobilization
           data and analyses are reliable.

           We identified, based on congressional interest and our knowledge
           of DOD issues, selected demographic and deployment variables to
           review. We then worked with the Defense Manpower Data Center
           (DMDC) to identify the data fields within DMDC's Contingency
           Tracking System (CTS) that best provided information about the
           selected demographic and deployment variables we wanted to
           analyze. Although we wanted to analyze the locations to which
           reservists were deployed and the units with which reservists were
           deployed, DMDC officials said, and we agreed based on our review
           of the data, that the data were not reliable enough for those
           purposes. Our selected variables included

           o  the number of deployed reservists who volunteered for at least
           one deployment;
           o  the number of deployed reservists who have served one, two, or
           three or more deployments;
           o  the race and ethnicity of the deployed reservists;
           o  the gender of the deployed reservists;
           o  the state of residence of the deployed reservists;
           o  the number of deployed reservists who were Selected Reserve,
           Individual Ready Reserve, Standby Reserve, or Retired Reserve;
           o  the number of deployed reservists who were citizens at the time
           of their deployment;
           o  the number of days the reservists were deployed; and
           o  the top occupational areas for reservists deployed in support
           of GWOT.

           To address objective 1, we obtained and analyzed data for
           September 2001 through June 2006 from DMDC's CTS. CTS consists of
           two files-the activation file, which tracks activations and
           mobilizations, and the deployment file, which tracks deployments.
           Using CTS data from both files, we analyzed the number of National
           Guard and Reserve servicemembers mobilized and deployed in support
           of GWOT, as well as selected demographic and deployment variables,
           using statistical analysis software.

           To address objective 2, we performed a data reliability assessment
           on the data provided by DMDC from CTS' activation and deployment
           files. We requested DMDC reports that replicated our analyses and
           then compared those report results to our analyses, and we
           reviewed the programming code DMDC used to generate those reports.
           To assess the reliability of CTS data, we obtained an
           understanding of the data, the file structure, the sources of the
           data, and relevant DOD guidance.1 Specifically, we (1) performed
           electronic testing of the data files for completeness (that is,
           missing data), out-of-range values, and dates outside of valid
           time frames; (2) assessed the relationships among data elements
           (for example, determining whether deployment dates were
           overlapping since each record in the deployment file is intended
           to represent one deployment); (3) reviewed existing information
           about the data and the systems that produced them; (4) interviewed
           department officials to identify known problems or limitations in
           the data, as well as to understand the relationship between the
           two files and how data are received from the services, cleaned
           ("rebaselined"), and processed by DMDC; and (5) compared
           "prerebaselined" mobilization data to "postrebaselined"
           mobilization data to determine the extent to which the data
           changed as a result of the cleaning effort.2 When we found
           discrepancies (for example, overlapping deployment dates), we
           worked with DMDC to understand the discrepancies.

           In our interviews with DMDC officials, we discussed the purpose
           and uses of CTS, the service data rebaselining effort and the
           internal controls for verifying data analyses, monthly validation
           of data, and performing data analyses. Similarly, we discussed
           data collection, processing, and reliability issues as well as
           service-specific data issues and the rebaselining effort with
           officials from the Office of the Assistant Secretary of Defense
           for Reserve Affairs and from each of the reserve components,
           including the U.S. Army National Guard, the U.S. Air National
           Guard, the U.S. Army Reserve, the U.S. Navy Reserve, the U.S.
           Marine Corps Reserve, and the U.S. Air Force Reserve. We also
           discussed the reliability of the services' data, the rebaselining
           effort, and the results of a previous Joint Staff review of the
           quality of service data within CTS with officials in the Joint
           Chiefs of Staff Manpower and Personnel office. Finally, we
           interviewed officials from the Deployment Health Surveillance
           Directorate and the Army Medical Surveillance Activity about the
           quality of the deployment data and how they use the data.

           In the course of our review, we determined that some data fields
           were highly unreliable. For example, electronic testing indicated
           that data on location and reservist unit information were missing
           in many cases. Based on our conversations with DMDC and our
           understanding of the data system, we decided not to conduct lower
           level analyses (for example, analyses of reservists' assigned
           units) because the results would be less reliable than aggregate
           level analyses. Although we are reasonably confident in the
           reliability of most CTS data fields at the aggregate level,
           because we could not compare source documentation from each of the
           services to a sample of DMDC data, we could not estimate precise
           margins of error. Consequently, we used the data for descriptive
           purposes, and we did not base any recommendations on the results
           of our analyses. In addition, we presented only higher level,
           aggregate data from fields that we determined were sufficiently
           reliable for our reporting purposes. For these purposes, and
           presented in this way, the CTS data we use are sufficiently
           reliable with the following caveat: The Army had not completed its
           rebaselining effort for mobilization data before the completion of
           our review, and we could not, therefore, assess the reliability of
           Army mobilization data to the same extent as those of the other
           services. However, based on our electronic testing, data
           comparisons, and interviews with officials, we believe that the
           data are sufficiently reliable to present as descriptive
           information.

           To assess the reliability of DMDC's reports (that is, its own
           analyses) of CTS data, we compared our independent analyses of
           National Guard and Reserve servicemembers' mobilization and
           deployment statistics with results that DMDC provided from its own
           analyses of the same data. To pinpoint differences in analytical
           assumptions, we reviewed the statistical code DMDC used to produce
           its reports and compared it with our programming code. Through an
           iterative process, we noted errors in DMDC's programs and
           requested changes and reruns of the data. We worked with DMDC to
           ensure that discrepancies were not caused by differences in our
           analytical assumptions. Where there were discrepancies, we reached
           the following consensus on how to address them:

           o  Removed the Coast Guard entries from our analyses of the CTS
           database since, as we state in this report, the Coast Guard
           Reserve is under the day-to-day control of the Department of
           Homeland Security rather than DOD.
           o  Combined a reservist's Social Security number with his or her
           reserve component to create a unique identifier. DMDC officials
           said they do this because they are unsure where the source of the
           error is when they find that a Social Security number corresponds
           with two reserve components for a deployment during approximately
           the same time period. DOD's policy,3 when there is a duplicate
           Social Security number for more than one reserve component, is to
           count both transactions. However, the use of duplicate Social
           Security numbers results in overcounting. Specifically, the June
           2006 file had 38 reservists with overlapping mobilizations, 20
           reservists with overlapping deployments, and more than 800
           deployed reservists who appeared to have legitimately changed
           components. To compensate for the 58 "errors" where DMDC did not
           know which mobilization or deployment to count, it double-counted
           all 58 reservists. Likewise, the 800 deployed reservists who
           changed reserve components during GWOT were also double-counted.
           o  Removed reservists from all analyses when their reserve
           component category is unknown, so that the numeric totals across
           analyses would be consistent. DMDC officials said that this is an
           undocumented standard operating procedure.
           o  Utilized the reservists' information for most recent deployment
           to provide the most current information possible in cases where a
           reservist deployed more than once.
           o  Calculated the length of a reservist's deployment by including
           both the day the deployment began and the day on which the
           deployment ended. Thus, the number of days deployed is inclusive
           of the beginning and end dates.
           o  Combined the race categories for Asian, Native Hawaiian, and
           Other Pacific Islander because, prior to 2003, the distinction
           between these two groups was not captured in the data.

           After clarifying and agreeing on the analytical assumptions, we
           again reviewed DMDC's code and compared its results with our own
           to determine whether and why there were remaining discrepancies.
           We also requested written documentation of DMDC's internal control
           procedures for the CTS data and, when no documentation was
           available, interviewed knowledgeable officials about existing
           internal control procedures. Using the framework of standards for
           internal control for the federal government,4 we compared the
           information from those documents and interviews with our numerous,
           iterative reviews of DMDC's statistical programs used to generate
           comparative reports to assess the reliability of DMDC-generated
           reports from CTS. We determined that the reports DMDC generated
           for our review were not sufficiently reliable for our reporting
           purpose. Thus, we completed our own data analyses.

           We performed our work from December 2005 through August 2006 in
           accordance with generally accepted government auditing standards.

           Appendix II: Data on Reservists Deployed in Support of GWOT through June 30, 2006
			  
			  Our analysis of DOD data indicates that most reservists who
           deployed in support of GWOT through June 30, 2006, were part of
           the Selected Reserve (see table 5). In addition, California,
           Texas, Pennsylvania, and Florida had the highest numbers of
           reservists who have deployed in support of GWOT through June 30,
           2006 (see table 6).

Table 5: Reserve Component Categories for Reservists Deployed in Support
of the Global War on Terrorism by Fiscal Year through June 30, 2006

                                       2001a   2002    2003    2004    2005 2006b 
Army     Selected Reservec             3,647  6,920  40,355  47,869  60,172            23,765
National Individual Ready     0      2           17      48      49      50 
Guard    Reserve/Inactive                                                   
         National Guardd,                                                   
         e                                                                  
         Total            3,647  6,922       40,372  47,917  60,221  23,815 
Air      Selected Reserve              1,522 19,515  22,605  13,479  12,630             9,882
National                                                                    
Guard    Total            1,522 19,515       22,605  13,479  12,630   9,882 
Army     Selected Reserve              1,172  4,994  35,181  27,915  21,884            11,531
Reserve  Individual Ready     9    189          951   1,427   2,853   1,516 
         Reserve/Inactive                                                   
         National Guard                                                     
         Standby/Retired      0     22          113     132     140     133 
         Reservef, g                                                        
         Total            1,181  5,205       36,245  29,474  24,877  13,180 
Navy     Selected Reserve                217  2,506   5,438   5,348   5,110             5,658
Reserve  Individual Ready    27     73          212     163     305     177 
         Reserve/Inactive                                                   
         National Guard                                                     
         Standby/Retired      1      1            5       3       2       8 
         Reserve                                                            
         Total              245  2,580        5,655   5,514   5,417   5,843 
Marine   Selected Reserve                  8  1,169   9,568   5,593   3,052               864
Corps    Individual Ready     0    182          378     262     242      62 
Reserve  Reserve/Inactive                                                   
         National Guard                                                     
         Standby/Retired      0      0            6       4       5       3 
         Reserve                                                            
         Total                8  1,351        9,952   5,859   3,299     929 
Air      Selected Reserve                502 11,592  11,591   8,777   8,157             6,421
Force    Individual Ready     6     35          139     137     241      41 
Reserve  Reserve/Inactive                                                   
         National Guard                                                     
         Standby/Retired      0     10           73      22      15      14 
         Reserve                                                            
         Total              508 11,637       11,803   8,936   8,413   6,476 
Total                                  7,111 47,210 126,632 111,179 114,857            60,125

Source: GAO analysis of DOD data.

Notes: Title 10 U.S.C. 10145(b) provides that the Army National Guard and
the Air National Guard are to be in the Ready Reserve. Reservists can be
counted more than once if they deployed more than once in different fiscal
years.

aFiscal year 2001 data are for September 11, 2001, through September 30,
2001.

bFiscal year 2006 data are for October 1, 2005, through June 30, 2006.

cThe Selected Reserve, part of the Ready Reserve, is composed of those
units and individuals designated by their respective services and approved
by the Chairman of the Joint Chiefs of Staff as so essential to initial
wartime missions that they have priority for training, equipment, and
personnel over all other reserve elements.

dThe Individual Ready Reserve consists mainly of trained individuals who
have previously served in active component units or in the Selected
Reserve and who have a remaining military service obligation.

eThe Inactive National Guard consists of National Guard personnel who are
attached to a specific unit but are temporarily unable to participate in
regular training. Currently, only the Army National Guard uses this
category.

fThe Standby Reserve consists of personnel who have completed all
obligated or required service or have been removed from the Ready Reserve
because of civilian employment, temporary hardship, or disability.

gThe Retired Reserve consists of personnel who have been placed in a
retirement status based on the completion of 20 or more qualifying years
of active component or reserve component service.

Table 6: State, Territories, and Armed Forces Areas of Residence for
Reservists Deployed in Support of the Global War on Terrorism through June
30, 2006

Residence of         Army      Air                  Marine     Air         
deployed         National National    Army    Navy   Corps   Force 
populationa         Guard    Guard Reserve Reserve Reserve Reserve   Total
Alaska                471      979     147       4      27      42   1,670 
Alabama             5,301    1,380   2,484     387     332     596  10,480 
Arkansas            3,891      830   1,078      49     139     132   6,119 
Arizona             2,005      583   1,208     138     430     697   5,061 
California          6,811    1,882   5,910   3,769   3,321   2,811  24,504 
Colorado              939      889   1,441      27     223     759   4,278 
Connecticut         1,753      627     738     141     206     226   3,691 
District of           131       27     174     234       9      36     611 
Columbia                                                           
Delaware              502      386     252      10      39     438   1,627 
Florida             5,232      683   4,478   1,347     929   2,448  15,117 
Georgia             5,860    1,543   3,623     407     732     913  13,078 
Hawaii              2,058      453     856     137      63     202   3,769 
Iowa                3,963      941   1,431      19     153      33   6,540 
Idaho               2,004      710     329      11      54      50   3,158 
Illinois            4,032    1,910   3,272     103     683     454  10,454 
Indiana             4,328    1,188   1,583      60     331     685   8,175 
Kansas              2,394      978   1,760      87     100     287   5,606 
Kentucky            2,803      628   1,347      33     190      90   5,091 
Louisiana           5,877      416   1,271     112     435     613   8,724 
Massachusetts       2,965    1,006   1,684      76     357     758   6,846 
Maryland            1,839    1,195   2,215     584     432     487   6,752 
Maine               1,407      647     417      79      26      22   2,598 
Michigan            3,560    1,542   1,824      72     681     403   8,082 
Minnesota           4,699    1,209   2,064      46     126     592   8,736 
Missouri            4,199    1,032   2,605     110     262     580   8,788 
Mississippi         5,647    1,308   1,263     219     126     190   8,753 
Montana             1,302      582     472       5      35      25   2,421 
North Carolina      6,287      935   2,909     284     697     755  11,867 
North Dakota        1,627      167     173       2       4      28   2,001 
Nebraska            2,007      687   1,262      16      73      58   4,103 
New Hampshire       1,121      417     553      30      54      71   2,246 
New Jersey          2,234    1,332   1,506     215     479     839   6,605 
New Mexico            978      482     531      52      79     122   2,244 
Nevada                612      541     335      73     153     137   1,851 
New York            4,571    2,566   4,833     179   1,406     898  14,453 
Ohio                4,047    2,552   4,452     125     613   1,402  13,191 
Oklahoma            2,951    1,406   1,151      51     135     597   6,291 
Oregon              3,112      301     419      49     223     302   4,406 
Pennsylvania        6,053    2,368   5,415     269     828   1,322  16,255 
Rhode Island        1,092      474     253     109      61      14   2,003 
South Carolina      3,760      817   1,767     268     179     996   7,787 
South Dakota        1,916      571     246       1       2      23   2,759 
Tennessee           5,519    1,612   1,855     103     368      94   9,551 
Texas               6,901    1,503   7,271   1,233   1,475   2,421  20,804 
Utah                2,697      811     986      10     228     569   5,301 
Virginia            3,238      879   3,756   1,759     905     585  11,122 
Vermont             1,294      410     135       1      12      13   1,865 
Washington          4,320      839   1,808     225     383   1,274   8,849 
Wisconsin           4,177    1,291   2,257      39     282     603   8,649 
West Virginia       1,975      879   1,137       6      45      75   4,117 
Wyoming               694      407     115       4       9      16   1,245 
Territories and     4,366      670   3,390   4,701      15     243  13,385 
Armed Forces                                                       
areas                                                              
Unknown or Not         22      703     651   2,345     139     743   4,603 
applicable                                                         
Total             163,544   51,174  95,092  20,415  19,288  28,769 378,282 

Source: GAO analysis of DOD data.

Notes: Territories include the District of Columbia, American Samoa, the
Northern Mariana Islands, the Virgin Islands, Guam, and Puerto Rico. Armed
Forces areas include Armed Forces of the Americas, Armed Forces Europe,
and Armed Forces Pacific. This population represents 13,385, or about 4
percent, of the total number of deployed reservists.

aReservists who had more than one deployment are counted by the state of
residence of record for the most current deployment.

38 DOD Instruction 7730.54 and DOD Instruction 6490.03.

39 DOD Instruction 7730.54.

40 DOD Instruction 7730.54.

41 GAO/AIMD-00-21.3.1.

42 During the ongoing monthly validation process, DMDC officials said that
they apply a series of undocumented business rules to identify and address
inconsistencies in the data provided by the services and comparable data
reported in DEERS and by DFAS.

43 GAO, Government Auditing Standards, GAO-03-673G (Washington, D.C.: June
2003).

44 GAO-03-673G, sections 7.12 b, 7.31, and 7.59.

1 DOD Instruction 6490.03, DOD Instruction 1336.5, and DOD Instruction
7730.54.

2 To assess the extent the data changed, we first received and analyzed
data provided by DMDC in a December 2005 file about reservists deployed in
support of GWOT. While DOD's rebaselining effort was still being
completed, we received updated data from DMDC in a June 2006 file about
reservists deployed in support of GWOT through June 2006, which allowed us
to compare prerebaselined December 2005 data with postrebaselined December
2005 data. We compared these two data submissions and analyzed the extent
to which the data had changed during the rebaselining effort.

3 DOD Instruction 7730.54.

4 GAO/AIMD-00-21.3.1.

Appendix III: Comments from the Department of Defense

Appendix IV: GAO Contact and Staff Acknowledgments

                                  GAO Contact

Derek Stewart (202) 512-5559 or [email protected]

                                Acknowledgments

In addition to the contact named above, Cynthia Jackson, Assistant
Director; Crystal Bernard; Tina Kirschbaum; Marie A. Mak; Ricardo Marquez;
Julie Matta; Lynn Milan; Rebecca Shea; and Cheryl Weissman made key
contributions to this report.

(350682)

GAO's Mission

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.

                 Obtaining Copies of GAO Reports and Testimony

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

Order by Mail or Phone

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

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

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

             To Report Fraud, Waste, and Abuse in Federal Programs

Contact:

Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: [email protected]
Automated answering system: (800) 424-5454 or (202) 512-7470

                            Congressional Relations

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

                                 Public Affairs

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

www.gao.gov/cgi-bin/getrpt? GAO-06-1068 .

To view the full product, including the scope

and methodology, click on the link above.

For more information, contact Derek Stewart at (202) 512-5559 or
[email protected].

Highlights of GAO-06-1068 , a report to congressional committees

September 2006

MILITARY PERSONNEL

DOD and the Services Need to Take Additional Steps to Improve Mobilization
Data for the Reserve Components

GAO has previously reported on the Department of Defense's (DOD) ability
to track reservists deployed to the theater of operations and made
recommendations. Reliable mobilization and deployment data are critical
for making decisions about reserve force availability and medical
surveillance. Because of broad congressional interest, GAO initiated a
review under the Comptroller General's authority to conduct evaluations on
his own initiative to determine (1) what DOD data indicate are the number
of reservists mobilized and deployed in support of the Global War on
Terrorism (GWOT) and the selected demographic and deployment
characteristics of those deployed and (2) whether DOD's reserve deployment
and mobilization data and analyses are reliable. GAO analyzed data and
data analyses from DOD's Contingency Tracking System (CTS) and interviewed
agency officials.

What GAO Recommends

GAO is recommending that DOD standardize the use of key terms; provide
required data, such as location; and document its internal procedures,
processes, and assumptions for analyzing and verifying data analyses. DOD
generally concurred except for the need for DMDC to document its
assumptions. GAO believes that basic key assumptions should be documented.

GAO's analysis of DOD data indicates that more than 531,000 reservists
have been mobilized in support of GWOT as of June 30, 2006, and more than
378,000 reservists, or 71 percent of the number mobilized, have been
deployed. The number of reservists deployed increased through fiscal year
2003 and remained stable through fiscal year 2005. The majority of
reservists have been deployed once. GAO's analysis further indicates that
of the more than 378,000 reservists who have deployed in support of GWOT,
81 percent have spent a year or less deployed and 17 percent of reservists
have spent more than 1 year but less than 2 years deployed. Of those who
deployed, almost 98 percent were U.S. citizens. Since GWOT began, about 78
percent of reservists who were deployed were White, about 14 percent were
Black or African American, and almost 90 percent identified themselves as
non-Hispanic and 8 percent as Hispanic. Of those who were deployed, 89
percent were male and 11 percent were female. There were three
variables-volunteer status, location deployed, and unit deployed-required
by DOD policy for which the Defense Manpower Data Center (DMDC) could not
provide data because the data either did not exist or were not reliable
enough for the purposes of GAO's report.

GAO found the deployment and mobilization data used to be reliable for
providing descriptive information. However, the mobilization data, some
deployment data fields, and DMDC's processes for data analyses need
improvement. DMDC and the services have recently taken steps to improve
the reliability of mobilization data; however, additional steps are needed
to make mobilization data more reliable. DMDC and the services have
undertaken a large-scale, challenging effort to replace all previous
service-provided mobilization data in DMDC's CTS database with new data
from the services, referred to as "rebaselining." To date, the Air Force
has certified that it has rebaselined its data and Navy officials say they
have validated their personnel files and established a common baseline of
data with DMDC. The Army, which has mobilized the largest number of
reservists, has not completed its rebaselining effort and has not set a
deadline for completion. Also, DOD has not fully addressed other data
issues that could affect the accuracy and completeness of the data, such
as standardizing the use of key terms and ensuring that the services
address data issues identified by DMDC as well as provide data for all
required data fields, such as location, to DMDC. Also, because the data
analyses DMDC provided had numerous errors, GAO questions the
effectiveness of its verification procedures and other supporting
procedures, all of which DMDC has not documented. Until DOD addresses data
issues and DMDC documents the internal control procedures it uses to
analyze data and verify its analyses of the data, the information provided
to decision makers within Congress and DOD may be unreliable and decision
makers will not be in the best position to make informed decisions about
reserve force availability and reservists' exposure to health hazards.
*** End of document. ***