Military Personnel: DOD Needs More Data Before It Can Determine  
if Costly Changes to the Reserve Retirement System Are Warranted 
(15-SEP-04, GAO-04-1005).					 
                                                                 
Since September 11, 2001, National Guard and Reserve personnel	 
have served, and will continue to serve, a critical role in the  
Global War on Terrorism. As a consequence, Congress has focused  
attention on the fair treatment of reservists and directed GAO to
review, among other things, the Department of Defense's (DOD)	 
reserve retirement system. Currently, five legislative proposals 
exist that would lower the age when eligible reservists can	 
receive their annuities. GAO assessed (1) DOD's reserve 	 
retirement system as compared to certain aspects of the active	 
duty retirement system, (2) the extent to which DOD data suggest 
that changes to the reserve retirement system may be warranted,  
and (3) the potential costs and effects of the five legislative  
proposals.							 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-04-1005					        
    ACCNO:   A12378						        
  TITLE:     Military Personnel: DOD Needs More Data Before It Can    
Determine if Costly Changes to the Reserve Retirement System Are 
Warranted							 
     DATE:   09/15/2004 
  SUBJECT:   Data collection					 
	     Military personnel 				 
	     Military reserve personnel 			 
	     National Guard					 
	     Retirement benefits				 
	     Retirement pensions				 
	     Attrition rates					 
	     Policy evaluation					 
	     Cost estimates					 
	     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-04-1005

                 United States Government Accountability Office

GAO Report to Congressional Committees

September 2004

MILITARY PERSONNEL

  DOD Needs More Data Before It Can Determine if Costly Changes to the Reserve
                        Retirement System Are Warranted

                                       a

GAO-04-1005

[IMG]

September 2004

MILITARY PERSONNEL

DOD Needs More Data Before It Can Determine if Costly Changes to the Reserve
Retirement System Are Warranted

                                 What GAO Found

The active duty and reserve component retirement systems require 20 years
of service for vesting, provide annuities that are based on 2.5 percent of
basic pay for every creditable year served, and offer options to
participate in the government's Thrift Savings Plan. Active duty members
receive an immediate annuity upon retirement based on pay rates in effect
on the date of retirement. Eligible reservists must wait until their 60th
birthday to receive an annuity. The annuity of retired reservists who
remain subject to recall is based on pay rates in effect when they reach
their 60th birthday.

Current data do not provide DOD with the information it needs to determine
what changes, if any, to the reserve retirement system are needed. DOD has
not established thresholds for attrition attributable to retirement and,
without thresholds, it is not possible to assess the extent to which DOD
is retaining the desired numbers of senior reservists or the extent to
which retirement benefits influence decisions to leave or stay in the
reserves. Nevertheless, GAO's analysis of DOD data for selected periods
when reservists were deployed shows that the overall makeup of the reserve
force by years of service, age, and pay grade has remained relatively even
and that overall reserve retention rates, to date, have remained
relatively stable. DOD has, however, identified certain high-demand
occupational specialties, such as military police, where retention rates
have decreased. While these cases do not necessarily suggest that uniform
retirement reform is required, they do suggest that targeted corrective
actions of some sort may be needed.

The estimated cost of lowering the age of receipt, coupled with an
associated increase in retiree health care costs, is significant and
considering the fact that military basic pay will continue to grow and
health care costs will continue to rise, the longer-term cost will be
significant. Importantly, all reservists-not just the minority of
reservists who have deployed to recent operations-would be covered under
this change. In addition, reservists who are experiencing increased
deployments may not be the beneficiaries of a reduction in the age of
receipt, given that only one in four reservists currently stays in the
reserves long enough to retire. Finally, DOD has not studied changes to
the reserve retirement system in the context of the total force. Changes,
if not evaluated from a total force perspective, could result in a
significant increased cost with little or no benefit and may produce
unintended consequences on DOD's active duty forces.

Estimated 10-Year Increase in Appropriations of Five Legislative
Proposals, Fiscal Years 2005-2014 (dollars in billions)

Age 55 with

Increased Immediate Age 53 on a Age 55 on a reduced

costs annuity Age 55 sliding scale sliding scale annuity

Retirement fund $20.45 $7.60 $2.88 $2.88 $3.59

Health care $14.24 $6.03 $2.49 $2.39 NA

Total $34.69 $13.63 $5.37 $5.27 $3.59

Source: GAO analysis of DOD data.

United States Government Accountability Office

Contents

Letter

Results in Brief
Background
The Reserve Component Retirement System Mirrors the Active

Duty Retirement System but Has Differences The Extent to Which Changes Are
Warranted to the Reserve Retirement System Is Unknown Proposed Changes to
Retirement System Are Costly and Need to

Be Made in Context of the Total Force Conclusions Recommendations for
Executive Action Agency Comments

                                       1

                                      3 6

12

17

25 31 32 32

Appendix I Scope and Methodology

Appendix II	Key Provisions of the Reserve Retirement Points System

Appendix III Reserve Workforce Profile Data

Appendix IV Attrition Rates for Reservists

Appendix V Comments from the Department of Defense

                      Appendix VI Related GAO Products 54

Tables

Table 1: Availability of Retirement Provisions 12 Table 2: Estimated
10-Year Increase in Appropriations of Five Legislative Proposals between
Fiscal Year 2005 and 2014 28

Figures

Figure 1: Organizational Chart of the Ready Reserve 8

Figure 2: Percentage of Total Personnel, by Service, Serving on Active
Duty or in One of the Reserve Components in Fiscal Year 2003 10

Figure 3: Total DOD Reserve Component Retention Rates, and Retention Rates
for Officer and Enlisted Reservists, Fiscal Years 1991 through 2003 23

Figure 4: Reserve Enlisted Personnel Average Years of Service for Calendar
Years 1991, 1997, and 2003 41 Figure 5: Reserve Enlisted Average Age for
Calendar Years 1991, 1997, and 2003 42 Figure 6: Reserve Enlisted
Personnel Average Pay Grade for Calendar Years 1991, 1997, and 2003 43
Figure 7: Reserve Officer Average Years of Service for Calendar Years
1991, 1997, and 2003 44 Figure 8: Reserve Officer Average Age for Calendar
Years 1991, 1997, and 2003 45 Figure 9: Reserve Officer Average Pay Grade
for Calendar Years 1991, 1997, and 2003 46 Figure 10: Reserve Attrition by
Years of Service for Fiscal Year 1991 48 Figure 11: Reserve Attrition by
Years of Service for Fiscal Year 1997 49 Figure 12: Reserve Attrition by
Years of Service for Fiscal Year 2003 50

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 15, 2004

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

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

Over the past decade, the Department of Defense (DOD) has increasingly
relied on members of the National Guard and Reserve to carry out its
military operations domestically and abroad. At the end of fiscal year
2003,
DOD had 1.2 million reservists.1 Since September 11, 2001, DOD has called
more than 335,000 reservists to active duty, and reservists will continue
to
serve a large critical role in the Global War on Terrorism.

Congress has focused attention on the fair treatment of reserve
component personnel as compared to their active duty counterparts. For
example, in recent years Congress has improved health care benefits not
only for reservists but for their families as well.2 Currently, several
legislative proposals are under consideration to change the reserve
retirement benefit that would lower the age when eligible reservists can
receive their annuities.

1 For the purpose of this report, the term reserve component, or
reservists, refers to the collective forces of the Army Reserve, the Army
National Guard, the Naval Reserve, the Air Force Reserve, the Air National
Guard, and the Marine Corps Reserve.

2 See GAO, Military Personnel: Observations Related to Reserve
Compensation, Selective Reenlistment Bonuses, and Mail Delivery to
Deployed Troops, GAO-04-582T (Washington, D.C.: Mar. 24, 2004).

House Report 107-436 accompanying the Fiscal Year 2003 National Defense
Authorization Act3 directed us to review, among other things, DOD's
reserve component retirement system.4 We assessed (1) DOD's reserve
retirement system as compared with certain aspects of the active duty
retirement system, (2) the extent to which DOD data suggest that changes
to the reserve retirement system may be warranted from a work force
management perspective, and (3) the potential costs and effects of
legislative proposals that address reserve retirement reform.

A number of factors need to be considered when assessing the need for
reserve retirement reform. Among them are the provisions of the active
duty retirement system, the magnitude of any problemsfrom a
workforce management and retention perspectivethat changes to the
reserve retirement system might potentially foster, other factors such as
costs to fund the retirement system and the health care benefits afforded
to retired reservists,5 and the impact on the active duty forces. Any
proposed changes that would permanently expand benefits should also be
considered in the context of the serious fiscal challenges facing this
country. The long-term fiscal pressure created by the known demographic
trends and rising health care costs, combined with increased homeland
security and defense commitments, and moreover, the gap between the
estimated costs of DOD's transformation needs and likely budgetary
resources, sharpen the need to carefully scrutinize the longer-term costs
and implications before these changes are enacted into law. For fiscal
year 2004, DOD estimates that it will need approximately $15.1 billion to
fund its current reserve component retirement system.

To conduct our review, we reviewed DOD reserve retirement policies and
studies conducted by DOD and others that address reserve retirement

3 Public Law 107-314, Dec. 2, 2002.

4 This report is one of a series of GAO reports that address compensation
and benefits for reserve personnel in response to this mandate. See GAO,
Military Personnel: DOD Needs More Data to Address Financial and Health
Care Issues Affecting Reservists, GAO-03-1004 (Washington, D.C.: Sept. 10,
2003); Military Personnel: DOD Needs to Assess Certain Factors in
Determining Whether Hazardous Duty Pay Is Warranted for Duty in the Polar
Regions, GAO-03-554 (Washington, D.C.: Apr. 29, 2003); and Military
Personnel: Preliminary Observations Related to Income, Benefits, and
Employer Support for Reservists During Mobilizations, GAO-03-549T
(Washington, D.C.: Mar. 19, 2003).

5 Currently, retired reservists are eligible to participate in DOD's
managed health care system when they reach age 60. Four legislative
proposals would reduce the age of eligibility.

issues. We also collected and analyzed attrition data contained in DOD's
Reserve Components Common Personnel Data System as well as attrition and
cost data from the Office of the Assistant Secretary of Defense for
Reserve Affairs and the Office of the DOD Actuary. We reviewed three
points in time1991, 1997, and 2003because these three
points represent key periods when reservists were called to active duty to
support military operations. These included Operations Desert Shield/
Desert Storm, military operations in the Balkans and Southwest Asia, and
current operations in Iraq and Afghanistan. In addition, we analyzed the
results of recent DOD web-based career satisfaction surveys that were
administered to reservists in each of the components. Based on our review
of databases we used, we determined that the DOD-provided data were
reliable for our purposes. Finally, we interviewed officials in the Office
of the Assistant Secretary of Defense for Reserve Affairs, the Office of
the DOD Actuary, the DOD Office of the Assistant Secretary for Health
Affairs, and DOD's six reserve components the Army Reserve, the
Army National Guard, the Naval Reserve, the Air Force Reserve, the Air
National Guard, and the Marine Corps Reserve. We conducted our review from
November 2003 through July 2004 in accordance with generally accepted
government auditing standards. Further information on our scope and
methodology appears in appendix I.

The reserve component retirement system encompasses many of the aspects of
the active duty retirement system, yet each system has certain unique
provisions. For example, both systems require a minimum of 20 years of
service for vesting and eligibility for retirement benefits, provide
retirement annuities that are based on 2.5 percent of basic pay for every
creditable year served, and offer opportunities to participate in the
federal government's Thrift Savings Plan. The primary differences between
the two systems are the age when active duty and reserve personnel receive
their annuities and the effective pay rates that are used to calculate
annuities. Eligible active duty military members receive an immediate
annuity upon retirement that is based on pay rates in effect on the date
of retirement. Eligible reservists must wait until age 60 to receive an
annuity. The annuities of retired reservists who remain subject to recall
are based on pay rates that are in effect when reservists reach their 60th
birthday and, accordingly, their retirement is calculated at these higher
pay rates.

Current available data do not provide DOD with the information it needs to
determine what changes, if any, to the reserve retirement system are
warranted. While DOD has established enlisted personnel attrition ceilings
as a performance metric for each of its six reserve components, it has not

  Results in Brief

established attrition ceilings for its reserve officers. In addition, it
has not established thresholds for attrition attributable to retirement.
For example, almost 18,000 reservists, representing 2.1 percent of DOD's
selected reservists, retired from reserve duty in fiscal year 2003.
Without thresholds or targets, however, it is not possible to assess the
extent to which DOD is retaining the desired numbers of senior military
members or to what extent retirement provisions influence decisions to
leave or stay in the reserves. Furthermore, some retention rates may be
artificially high because, at various times since September 11, 2001, DOD
has implemented "stop loss" policies that have precluded
servicemembers-both active and reserve-from separating from military
service. Of all the services, the Army has made the greatest use of the
stop loss policy. Nevertheless, our analysis of various DOD data for
selected periods of time when reservists were called to active duty to
support military operations reveals that the overall composition of the
reserve force by years of service, age, and pay grade has remained
relatively even and that DOD has not experienced significant declines in
any of these variables. Similarly, departmentwide reserve retention rates
for fiscal years 1991 through 2003 remained relatively stable, fluctuating
between 76.5 and 82.9 percent, and the reserve components generally met
their retention thresholds for enlisted personnel in recent years. In
fiscal year 2003, for example, all components met their thresholds, with
the exception of the Army National Guard and the Air National Guard who
missed their thresholds by less than 1 percent. DOD has, however,
identified certain high-demand occupational specialties where retention
rates have decreased. While these cases do not necessarily suggest that
uniform retirement reform is required, they do suggest that targeted
corrective actions of some sort may be needed. For example, the retention
rate for the military police within the Army Reserve dropped from
approximately 67 percent in fiscal year 2000 to 49 percent in fiscal year
2003. Similarly, retention rates decreased for Army National Guard members
in aviation-dropping from 80 percent to approximately 30 percent in fiscal
year 2002. Finally, recent responses by reservists to surveys administered
DOD-wide reflect these retention statistics. Responses to questions about
career intention indicated that most reservists intended to continue their
service in the reserves. For example, in the most recent survey
available-September 2003-67.3 percent of the reservists who responded to
the survey indicated that they would continue in the reserves and noted
that reserve retirement was one reason for joining the reserves. However,
only 52.6 percent of respondents in the Army National Guard and 53.5
percent of respondents in the Army Reserve, who were or had been called to
active duty, indicated that they would stay in the Guard or Reserves.

The estimated cost of current legislative proposals to change the reserve
component retirement system is significant and will have a number of
associated effects. First, the estimated cost of lowering the age of
receipt, from age 60 to as low as 53 with limitations-or, in some cases,
lower- will be significant, given that all reservists who qualify for
retirement-and not just the minority of reservists who have deployed to
recent contingency operations-would be covered under this change. Based on
DOD data, our cost estimates of current legislative proposals are
substantial, projecting that the cost of reserve retirement will increase
from about $3 billion to over $20 billion, depending on the proposal, over
the next 10 years. The Congressional Budget Office had not estimated the
cost of the retirement benefit for all five legislative proposals at the
time of our review. When associated health care costs are factored in, the
total projected cost increase ranges from over $3 billion to almost $35
billion. Considering the fact that military basic pay will continue to
increase and health care costs will continue to rise, the longer-term cost
will continue to be significant. Second, lowering the age of receipt will
align the reserve retirement system more closely with the active duty
retirement system, but this change may not affect reservists who are
experiencing increased deployments and associated hardships. Many of these
reservists are not likely to be the beneficiaries of a reduction in the
age of receipt of annuity, given that only one in four reservists
currently stays in the reserves long enough to become eligible for
retirement. Third, retirement is only one human capital management tool
available to DOD's workforce managers to compensate servicemembers for
hardships experienced during their careers. For example, DOD has several
special pays and allowances it can use to compensate deployed
servicemembers, such as hazardous duty pay and family separation
allowances. In addition, Congress has approved, but DOD has yet to
implement, a special allowance for military personnel who experience
frequent and/or long duration deployments. Fourth, DOD has recently
identified a need to rebalance the reserve and active duty forces to
reduce the stress on certain high-demand occupations. Better use of these
human capital management tools can help DOD address the issues of equity
and the increased use of reservists in military operations. Finally, DOD
has not yet studied changes to the reserve retirement system in the
context of the total force, even though these changes could have
unintended consequences on DOD's active duty forces. For example, if the
reserve retirement system were changed to offer an immediate annuity after
20 years of creditable service, some personnel who have made career
decisions to remain in active duty until retirement might be inclined to
leave active duty and apply to serve their remaining time in a reserve
component.

This report contains recommendations that DOD (l) specify desired metrics
for measuring the attrition rates of senior officer and enlisted
reservists who are approaching retirement eligibility and, therefore, are
most likely to be affected by changes to DOD's reserve retirement system;
(2) determine if gaps exist between the desired and actual rates of
attrition; (3) identify changes, if any, to the current reserve retirement
system that would address these gaps, to the extent that they exist; and
(4) evaluate any changes to the reserve retirement system in the context
of the total force. In written comments on a draft of this report, DOD
stated that it concurred with each of the four recommendations.

Background 	DOD has relied on its reserve forces to varying degrees over
time. Reserve personnel are eligible to receive compensation the same as
that of active duty personnel when called to active duty.

Dependence on Reserves in Military Operations

Since the end of the Cold War, there has been a shift in the way DOD has
used its reserve forces. Previously, reservists were viewed primarily as
an expansion force that would supplement active forces during a major war.
DOD has since moved occupational specialties that are key to conducting
any military operation into the reserve forces and, therefore,
increasingly relies on reservists not only to supplement but also to
replace active forces in military operations worldwide. Accordingly, DOD
has called more than 335,000 reservists to active duty since September 11,
2001. Today's increased reliance on the reserve components is a product of
a number of factors: evolving total force policies, effects of force
downsizing, increasing mission demands, and the Abrams Doctrine. In the
aftermath of the Vietnam War, former Chief of Staff of the Army General
Creighton Abrams asserted his belief that U.S. armed forces should not go
to war again without the involvement of the National Guard and Reserve,
given that decisions to employ reservists would reflect the national
intent of the public. This philosophy became known as the Abrams Doctrine,
and it has profoundly influenced DOD's national defense strategies and
plans. In fact, and as pointed out in the Fiscal Year 2002 Annual Report
of the Reserve Forces Policy Board,6 DOD entered into the Global War on
Terrorism dependent on the reserve components as the cumulative effect of
30 years of adherence to the Abrams Doctrine. In the Global War on
Terrorism, the

6 "Current and Future Challenges." Reserve Component Programs: The Annual
Report of the Reserve Forces Policy Board, Office of the Secretary of
Defense, April 2003.

reserve components are considered full partners, providing nearly one-half
of the personnel who currently comprise the total force. The Reserve
Forces Policy Board cautioned that if DOD retains this doctrine, reserve
forces will be committed for the duration of military operations and used
and stressed as never before, adding that appropriate adjustments will be
necessary to spread the burden and provide for the long-term
sustainability of the force.

Composition of the Reserve Forces

At the end of fiscal year 2003, DOD had approximately 1.2 million Guard
and Reserve members in the Ready Reserve. Within the Ready Reserve, there
are three subcategories, the Selected Reserve, the Individual Ready
Reserve,7 and the Inactive National Guard.8 Figure 1 shows how the Ready
Reserve is organized.

7 The Individual Ready Reserve is comprised principally of individuals who
(1) have had training, (2) have served previously in an active component
or the Selected Reserve, and (3) have some period of their military
service obligation remaining. Members may voluntarily participate in
training for retirement points and promotion with or without pay. Within
the Individual Ready Reserve, there is a special category of members who
have volunteered to be called to active duty under 10 U.S.C. 12304 when
needed.

8 The Inactive National Guard consists of National Guard personnel in an
inactive status in the Ready Reserve who are attached to a specific
National Guard unit. To remain in the Inactive National Guard, members
must muster once a year with their assigned unit, but they do not
participate in training activities. Inactive National Guard members may
not train for points or pay and are not eligible for promotion.

              Figure 1: Organizational Chart of the Ready Reserve

Source: GAO presentation of DOD data.

The Selected Reserve consists of reservists who serve in Selected Reserve
units, as Individual Mobilization Augmentees,9 as active Guard and Reserve
members,10 and as full-time support assets, such as military
technicians.11

DOD's selected reservists serve in one of six reserve components: the Army
Reserve, the Army National Guard, the Naval Reserve, the Air Force
Reserve, the Air National Guard, and the Marine Corps Reserve. The Army
National Guard and the Air National Guard comprise what is known as the
National Guard. As of fiscal year 2003, the Selected Reserve had a total
number of 875,072 members. All selected reservists are subject to being

9 Individual Mobilization Augmentees are trained individuals assigned to
an active component, Selective Service System, or Federal Emergency
Management Agency organization's billet, which must be filled on or
shortly after mobilization. Individual Mobilization Augmentees participate
in training activities on a part-time basis with an active component unit
in preparation for recall in a mobilization. The inactive duty training
requirement for Individual Mobilization Augmentees is decided by DOD
component policy and can vary from 0 to 48 drills a year. A minimum of 12
days of annual training is required of all Individual Mobilization
Augmentees.

10 Active Guard and Reserve personnel are members of the Selected Reserve
who are ordered to active duty or full-time National Guard duty for the
purpose of organizing, administering, recruiting, instructing, or training
the reserve component units. All Active Guard and Reserve members must be
assigned against an authorized mobilization position in the unit they
support. Units include the Navy Training and Administration of Reserve,
and the Marine Corps Active Reserve programs.

11 Military technicians-who provide full-time support to the reserve
components- concurrently serve as civilian employees and selected
reservists and receive retirement annuities from both the federal civilian
retirement system and the military reserve retirement system.

called to active duty before reservists in any of the other reserve
categories. The President has the authority to order up to 200,000
selected reservists to active duty involuntarily for any operational
mission-for not more than 270 days-without declaring a national emergency.
Therefore, reservists must be prepared to report for active duty within 24
hours.

The purpose of each selected reserve component is to provide trained units
and qualified persons available for active duty in the armed forces in
times of war or national emergency, and at such other times as the
national security may require, to fill the needs of the armed forces
whenever needed.12 The National Guard is unique in that it has dual
missions, both federal and state; when not in federal service, the Guard
is available for use as provided by the constitution and laws of the
state. The Guard is the only military force immediately available to a
governor in times of emergency, including civil unrest and natural or
manmade disasters. Under state law, the Guard provides protection of life
and property and preserves peace, order, and public safety.

Reliance on Guard and Reserve members varies from service to service.
Figure 2 shows that, in fiscal year 2003, 53 percent of the Army's
soldiers served in the Army National Guard or the Army Reserve and that 33
percent of Air Force military personnel served in the Air National Guard
and the Air Force Reserve. Figure 2 also shows that the percentage of
personnel serving in the Naval and Marine Corps Reserves was less than 20
percent for both components.

12 10 U.S.C. 10102.

Figure 2: Percentage of Total Personnel, by Service, Serving on Active
Duty or in One of the Reserve Components in Fiscal Year 2003

Percentage of personnel serving in active duty, guard, or reserve status

In 1948, Congress passed the Army and Air Force Vitalization and
Retirement Equalization Act-which established the reserve retirement
system-because it believed that reservists did not receive sufficient
salary as an incentive for a career and that, as reservists approached
mid-career, they would leave reserve service unless an additional
incentive were offered to them. Congress established a retirement annuity,
receivable at age 60 upon completion of at least 20 years of service, to
provide them with this additional incentive. The sole purpose of reserve
retirement is to provide a monetary incentive for qualified reserve
personnel to retain membership in the reserve components and continue
their training. This incentive provides the reserves with a pool of
skilled, trained, and readily available personnel to augment active duty
forces in times of national emergency. The reserve retirement system has
remained essentially unaltered since enactment in 1948. The Assistant
Secretary of Defense for Reserve Affairs, under the Office of the Under
Secretary of Defense for Personnel and Readiness, is responsible for DOD's
reserve

                                     100 90

                                       80

                                       70

                                       60

                                       50

                                       40

                                       30

                                       20

                                      10 0

Army Air Force	Marine Corps

                                      Navy

Army/Air Force/Marine Corps/Naval Reserve Army National Guard/Air National
Guard Active duty personnel Source: GAO analysis of DOD data.

                                Pay and Benefits

retirement policy, and the reserve retirement system is funded through
DOD's reserve military personnel appropriations.

Reservists are currently entitled to a range of pay and benefits,
depending on whether they are serving in part-time drilling or activated
full-time duty status. When in normal part-time drilling status,
reservists receive basic pay13 on a prorated basis and are entitled to
such benefits as unlimited access to commissaries, certain health care
benefits, educational benefits, and life insurance. Reservists activated
for contingency operations are eligible to receive the same compensation
and benefits as active duty personnel.14 While serving on active duty,
reservists are entitled to basic pay and special pays and allowances at
equivalent rates for active duty personnel and full health care benefits
for themselves and their dependents.

By law,15 DOD is required to undertake a comprehensive review of military
compensation every 4 years, and DOD compiles the findings and
recommendations of these reviews into a report known as the Quadrennial
Review of Military Compensation. In 1986, DOD was tasked by the President
to perform an evaluation of the benefits and costs of reserve
compensation. The resulting Sixth Quadrennial Review of Military
Compensation was the first comprehensive review of reserve compensation
that employed analytical models to assess the effects of the existing
reserve compensation system and evaluate the costs and benefits of
alternatives. The most recent quadrennial review assessed the
effectiveness of current military compensation policies in recruiting and
retaining a high-quality force. The Ninth Quadrennial Review of Military
Compensation was conducted in 2002.

13 Basic military compensation consists of basic pay, basic allowance for
housing, basic allowance for subsistence, and the federal tax advantage.
The federal tax advantage is to account for the tax-free status of housing
and subsistence allowances. It is the added amount of taxable income that
servicemembers would have to receive in cash if housing and subsistence
allowances were to become taxable in order for them to be as well off in
after-tax income as they are under the existing system. Basic compensation
does not include special and incentive pays and other allowances, such as
deployment pays.

14 For a listing of benefits, see Military Personnel: Active Duty Benefits
Reflect Changing Demographics, but Opportunities Exist to Improve,
GAO-02-935 (Washington, D.C.: Sept. 18, 2002).

15 37 U.S.C. 1008(b).

  The Reserve Component Retirement System Mirrors the Active Duty Retirement
  System but Has Differences

Both active duty and reserve component retirement systems operate under
the same military compensation principles. Differences exist between them
because the reserve retirement system was not intended to be comparable to
active duty retirement when it was created in 1948. Similarities between
the systems include a minimum of 20 years of service for vesting and
eligibility for retirement benefits, a calculation of retirement pay at
2.5 percent of basic pay for every creditable year served, and an
opportunity to participate in the federal government's Thrift Savings
Plan. The primary differences between the two systems are (1) the age when
active duty and reserve personnel receive their annuities and (2) the
effective pay rates that are used to calculate annuities. An eligible
active duty military member receives an immediate annuity upon retirement
that is based on pay rates in effect on the date of retirement. An
eligible reservist must wait until age 60-the age at which federal
civilian employees could retire in 1948-to receive an annuity. Retired
reservists who remain subject to recall receive an annuity that is based
on pay rates in effect when they reach age 60, and, accordingly, the
retirement is calculated at these higher pay rates. Table 1 shows various
retirement provisions that are available to active duty and reserve
personnel.

                 Table 1: Availability of Retirement Provisions

                                                                             Annuity  
                                     Annuity                       Annuity   based on 
                            Thrift   based on                     based on   higher   
           Annuity         Savings     2.5                        effective effective 
                                     percent                                
            after  Thrift    Plan   of         Receipt of Receipt basic pay basic pay 
                                    creditable            of                
Retirement   20    Savings matching  years of  annuity    annuity rates at  rates at  
            years                              upon       at      date      
provisions      of  Plan    funds    service   retirement age 60  of         age 60   
           service                                                discharge 
  Active                                                                    
duty                                                                     
personnel     X       X       Xa        X          X                  X     
 Reserve                                                                    
personnel     X       X                 X                    X       Xb             X 

Source: GAO.

aDOD has the authority to offer matching funds to active duty members
serving in critical occupations but had not exercised this authority as of
July 2004. See 37 U.S.C. 211(d).

bAn option exists for reservists to separate entirely from the reserves
upon completing 20 years of qualifying service. Those reservists who
select this option are no longer subject to being called to active duty.
However, their retirement annuities, which they do not receive until age
60, are based on pay rates that were in effect at the date of separation
from the service.

The Active Duty and Reserve Component Retirement Systems Were Not Intended
to Be the Same

The active duty and reserve component retirement systems contain many of
the same aspects, but they were not intended to be the same. Similarities
between the retirement systems are due to the fact that the underlying
principles for both retirement systems are the same. The purpose of the
military compensation system-of which retirement is one part-is to
attract, retain, and motivate the number of quality personnel needed to
maintain the desired level of national security. DOD has said that the
military retirement systems-both active duty and reserve- should support
and complement force management requirements (e.g., youth, vigor, and
career development opportunities); be integrated into the military
compensation system; and be structured to replace and maintain income for
retirees. Both retirement systems have the shared purpose of maintaining a
pool of skilled, trained, and readily available personnel subject to
recall during time of war or national emergency.

Differences between the retirement systems exist, however, because the
reserve retirement system was originally intended to compensate civilian
reserve retirees differently than active duty retirees. The active duty
retirement system, which was created in the mid-1800s, was originally
intended to provide for the separation of older officers with over 40
years of service to obtain a youthful and vigorous officer force, and
improve promotion opportunities for younger officers. This active duty
retirement system has been substantially modified over time. For example,
the provision enabling active duty military personnel to retire after 20
years was first introduced in 1915. More than 30 years later, the 20-year
provision was adopted for all active duty officers and enlisted personnel
in all services. When the reserve component retirement system was created
in 1948, Congress and the War and Navy Departments did not intend that it
should be comparable to the active duty retirement system. In contrast,
the purpose of the reserve retirement system was to improve retention, as
opposed to encouraging attrition, by providing a financial incentive for
qualified personnel to maintain their reserve affiliation and continue
training in the reserves. Establishing age 60-the age at which federal
civilian employees could retire in 1948-as the age at which reservists
would be eligible to receive retirement annuities provided a mechanism to
retain reservists for a longer period of time while compensating civilian
reserve retirees differently than active duty retirees.16

16 In 1967, Public Law 90-83, Sec. 1(75) created an entitlement for
federal civilian employees to receive an annuity at age 55 with 30 years
of service or at age 60 with 20 years of service.

The Active Duty and Reserve Component Retirement Systems Have Similar
Features

Both the active duty and the reserve component retirement systems require
20 years of service for vesting and benefit eligibility. The fact that
neither the active duty nor the reserve component retirement system
provides for reduced annuities if personnel serve less than 20 years
provides a strong incentive for servicemembers to either leave the
military after serving only a few years or stay in the military for 20
years or longer. For active duty members, each year of service counts as a
qualifying year for retirement purposes. Reservists must earn a minimum of
50 points, each year, for the year to qualify for retirement purposes.
Reservists typically train one weekend each month and earn 4 points for
each weekend (two drills per day). In addition, they earn 1 point for each
day of annual active duty training (typically 12 points for a 2-week
period) and an additional 15 points annually for being a member of a
reserve unit. Reservists earn an additional point for each day that they
are called to serve on active duty. Between fiscal year 1992 and fiscal
year 2001, reservists earned, on average, 64 retirement points per year.
In fiscal year 2003, reservists earned, on average, 138 retirement points.
The increase is likely due to the number of reservists who were called to
active duty. Appendix II provides a more detailed discussion of the
reserve retirement points system.

Generally speaking, both retirement systems apply a rate of 2.5 percent to
the number of creditable years served and then apply this rate to
servicemembers' basic pay to determine retirement annuities.17 In the case
of reserve personnel, the number of creditable years served is,
essentially, a prorated number of calendar years of service.
(Specifically, reserve years of creditable service are calculated by
dividing a reservist's accumulated retirement points by 360-roughly, the
number of days in a year.) Typically, current active duty servicemembers
who retire after serving 20 years on active duty receive retirement
annuities that are equivalent to 50 percent of average basic pay over
their last 3 years of service. Current active duty servicemembers who
retire after serving 30 years on active duty receive retirement annuities
that are equivalent to 75 percent of their

17 For active duty servicemembers who entered the service prior to
September 8, 1980, this rate is applied to a servicemember's final basic
pay at the date of retirement. For most reservists who entered the service
prior to September 8, 1980, this factor is applied to basic pay rates in
effect on the retired reservist's 60th birthday. For active duty
servicemembers who entered the service on or after September 8, 1980, this
factor is applied to a servicemember's average basic pay during the last
36 months of service. For most reservists who entered the service on or
after September 8, 1980, this factor is applied to the average annual
basic pay rates in effect during the 3 years prior to the reservist's 60th
birthday.

final basic pay.18 In contrast, current reserve members who retire after
serving all 20 years in the reserves, and who earn only the minimum 50
required points in each of those years, would receive annuities that will
be equivalent to approximately 7 percent of basic pay. Most reservists,
however, earn more than 50 points each year. Reservists can apply a
maximum of 90 points each year toward their retirement calculation- earned
from their weekend drills and annual 2-week period of active duty
training-but can apply up to 365 points each year if they are called to
active duty. Reservists who are called to active duty will earn more
points, which will translate into a higher retirement annuity, if they
serve their full 20 years. In addition, many reservists previously served
some time on active duty before affiliating with one of the reserve
components. Therefore, while their retirement annuities will be
significantly less than the annuities of their active duty counterparts,
the annuities of retired reservists who served on active duty at points
during their careers will be greater than 7 percent of basic pay.

Beginning in fiscal year 2001, both active duty and reserve personnel
became eligible to participate in the federal government's Thrift Savings
Plan on a voluntary basis. The Thrift Savings Plan is a tax-deferred
savings and investment plan only available to currently employed federal
government employees and provides active duty and reserve servicemembers
the opportunity to contribute additional funds to their retirement.
Service secretaries have the legislative authority, if they choose, to
provide matching funds for active duty personnel in critical military
occupational specialties, but not for reservists. Agency officials
frequently cite monetary recruitment and retention incentives as effective
for hard-to-fill positions.19 To date, the service secretaries have not
used this human capital flexibility for any active duty military
personnel.

18 An active duty servicemember has the option of retiring under an
alternative retirement system, known as REDUX, which reduces the
percentage factor of 2.5 percent-and, therefore, the amount of the
annuity-for each year that a servicemember retires with less than 30 years
of service. However, this alternative retirement system also provides a
servicemember with a $30,000 bonus at 15 years of service. The system,
which was made voluntary in fiscal year 2000, was designed to encourage
longer-term retention by providing an incentive for servicemembers who
complete 20 years of service, and who are otherwise eligible, to remain on
active duty beyond 20 years.

19 GAO, Human Capital: Effective Use of Flexibilities Can Assist Agencies
in Managing Their Workforces, GAO-03-2 (Washington, D.C.: Dec. 6, 2002).

The Main Differences between the Two Systems Are the Age of Receipt of
Annuity and the Effective Pay Rates Used to Calculate Annuities

One significant difference between the active duty and reserve retirement
systems is the age at which retirees receive their annuities. Active duty
personnel immediately qualify for receipt of an annuity after 20 years of
service, should they decide to retire, that compensates them for lost
income potential. The Defense Science Board Task Force on Human Resources
Strategy stated in a 2000 report that many, if not most, active duty
retirees suffer a "second-career" earnings loss when transitioning to the
civilian sector in their 40s or 50s and that immediate receipt of
retirement pay reduces the financial penalty associated with the
transition to a second career. As a result, the immediate annuity provides
a strong incentive for active duty personnel to retire when eligible.
Reservists similarly become eligible for retirement upon completing 20
years of service, but they are not eligible to receive an annuity until
their 60th birthdaythe age at which federal civilian employees
could retire when the reserve retirement system was enacted. The incentive
for reservists to separate once retirement eligibility is achieved is not
as strong, given that their retirement annuity is delayed until they reach
age 60.

A second significant difference between the two retirement systems is the
basic pay rate that is used to calculate retirement annuities. Active duty
retirement pay is based on basic pay rates that are in effect when active
duty servicemembers retire. Reserve retirement pay is calculated
differently, depending on retirement options that reservists select.
Reservists who attain 20 years of qualifying service for retirement
purposes and who elect to retire can, if otherwise eligible, (1) continue
their affiliation in the reserves in a nondrilling status, (2) transfer to
retired reserve status in which they retain their reserve affiliation but
are not required to train, or (3) request a discharge. Reservists who
continue their reserve affiliation in a nondrilling status do not drill or
receive basic pay, but they can acquire additional retirement points.20 In
addition, they remain subject to being involuntarily called back to active
duty. Reservists who transfer to retired reserve status do not earn basic
pay or additional retirement points. They are, however, also subject to
being involuntarily called back to active duty. Retirement pay for
reservists who retain some form of affiliation with the reserves, and who
are subject to involuntary recall to active duty until they reach age 60,
is based on pay rates that are in effect on the reservists' 60th birthday.
Accordingly, this provision

20 For example, reservists who are in a nondrilling status can earn
additional retirement points by successfully completing correspondence
courses.

  The Extent to Which Changes Are Warranted to the Reserve Retirement System Is
  Unknown

creates an incentive for reservists to retain their affiliation with the
reserves until age 60 in order to benefit from the calculation of their
annuity at the higher pay rate. Reservists who request a discharge from
the reserves once they become retirement eligible are not subject to being
called back to active duty. They also do not receive their annuities until
their 60th birthday, but their retirement annuities are derived from basic
pay rates in effect at the date of discharge.

Current available data do not provide DOD with the information it needs to
determine what changes, if any, to the reserve retirement system are
warranted. While DOD has established enlisted personnel attrition ceilings
as a performance metric for each of its six reserve components, it has not
established attrition ceilings for its reserve officers. In addition, it
has not established thresholds or targets for attrition attributable to
retirement. In fiscal year 2003, for example, 2.1 percent of DOD's
selected reservists retired from reserve duty. Without thresholds or
targets, it is not possible to assess the extent to which DOD is retaining
the desired numbers of senior military members or the extent to which
retirement provisions influence decisions to leave or stay in the
reserves. Furthermore, some retention rates may be artificially high
because, at various times since September 11, 2001, DOD has implemented
"stop loss" policies that have precluded servicemembers-both active and
reserve-from separating from military service. Of all the services, the
Army has made the greatest use of this policy. Nevertheless, our analysis
of various DOD data for selected periods of time when reservists were
called to active duty to support military operations reveals that the
overall composition of the reserve force by years of service, age, and pay
grade has remained relatively even and that DOD has not experienced
significant declines in any of these variables. Similarly, departmentwide
reserve retention rates for fiscal years 1991 through 2003 remained
relatively stable, fluctuating between 76.5 and 82.9 percent and the
reserve components generally met their retention thresholds for enlisted
personnel in recent years. In fiscal year 2003, for example, all
components met their thresholds, with the exception of the Army National
Guard and the Air National Guard who missed their thresholds by less than
1 percent. DOD has, however, identified certain high-demand occupational
specialties where retention rates have decreased. While these cases do not
necessarily suggest that uniform retirement reform is required, they do
suggest that targeted corrective actions of some sort may be needed. For
example, the retention rate for the military police within the Army
Reserve dropped from approximately 67 percent in fiscal year 2000 to 49
percent in fiscal year 2003. Similarly, the retention rate decreased for
Army National Guard

members in aviation, dropping from 80 percent to approximately 30 percent
in fiscal year 2002. Finally, recent responses by reservists to surveys
administered DOD-wide reflect these retention statistics. Responses to
questions about career intent indicated that most reservists intended to
continue their service in the reserves. For example, in the most recent
survey available-September 2003-67.3 percent of all reservists who
responded indicated that they would continue in the reserves and noted
that reserve retirement was a reason for joining the reserves.

Current Available Data Do Not Provide DOD with Complete Information

DOD does not have complete data to properly identify whether changes to
the reserve retirement system are needed. Our prior work21 has shown that
one of the critical success factors for strategic human capital management
is the use of reliable data to make human capital decisions. A fact based
performance-oriented approach to human capital management is crucial for
maximizing the value of human capital as well as managing relative risks.
Accordingly, DOD has established enlisted personnel attrition ceilings as
a performance metric, for each of its six reserve components, which vary
by component. For example, for fiscal years 2003 through 2005, the
attrition ceilings range from a low of 12 percent in the Air National
Guard to a high of 36 percent in the Naval Reserve. According to DOD,
these ceilings represent the maximum number of losses deemed acceptable in
a given fiscal year that are not to be exceeded. However, DOD has not
established attrition ceilings for its reserve officers.22 An official
within the Office of the Assistant Secretary of Defense for Reserve
Affairs stated that DOD historical data indicate that military officers
tend to be more career-focused than enlisted personnel. While on occasion
there are some officer retention issues, DOD has generally found that
long-term officer retention is more stable and higher than enlisted
personnel retention. Consequently, DOD has not identified a need for
establishing officer attrition ceilings.

21 GAO, Exposure Draft: A Model of Strategic Human Capital Management,
GAO-02-373SP (Washington, D.C.: March 2002).

22 DOD prefers to measure attrition rates rather than retention rates,
given that only a small portion of the reserve population is eligible for
reenlistment during any given year. In assessing enlisted personnel
retention trends in the reserve components, DOD computes enlisted
personnel attrition rates against established annual targets by dividing
total losses from a reserve component in a given fiscal year by the
average personnel strength for that component in the same year.

Furthermore, DOD has not established similar ceilings for the subset of
reserve attrition that can be attributed to retirement for officers or
enlisted personnel. For example, 17,959 reservists, representing 2.1
percent of the Selected Reserve, retired from selected reserve duty in
fiscal year 2003. Without ceilings or targets against which to measure
attrition attributable to retirement, however, it is not possible to
assess the extent to which DOD is retaining the desired number of senior
reserve officer and enlisted personnel or to determine whether changes to
the reserve retirement system are needed to manage the force. In other
words, without targets, DOD cannot determine if 2.1 percent attrition due
to retirement is desirable.

Identifying or establishing retirement goals would be beneficial to DOD
for several reasons. First, retirement is a force-shaping tool. As we
previously stated, the purpose of DOD's retirement systems-both active and
reserve-should support force management requirements (e.g., youth, vigor,
and career development opportunities) in addition to replacing or
maintaining income for retirees. Second, retirement policies provide the
primary financial incentive available to those senior personnel most
likely to be affected by retirement policies. According to DOD's Sixth
Quadrennial Review of Military Compensation, reserve retirement provisions
begin to shape the retention decisions of military personnel with as
little as 5 to 7 years of service, and the effect of these provisions on
retention decisions gradually increases as servicemembers approach 20
years of service. This can be attributed, in part, to the fact that DOD's
retirement systems do not provide reduced annuities to members who serve
less than 20 years. DOD officials have observed that the retirement
incentive has a noticeable retention effect that begins at the 12th to
14th year of service. DOD assumes that most personnel who have more than
14 years of service have decided to make a career out of the military and
will stay until they are eligible for retirement, and DOD data show that
attrition rates decline as servicemembers approach 20 years of service.
Accordingly, DOD does not offer retention bonuses to military members who
have served 14 years or longer. Third, retirement is a very expensive
force-shaping tool. As noted earlier, for fiscal year 2004, DOD estimates

that it will need about $15.1 billion to fund its current reserve
retirement

23

system.

Finally, DOD does not have the best data available to determine if changes
are needed because it has implemented "stop loss" policies in recent years
that have precluded servicemembers-both active and reserve-from separating
from military service. As a result, retention rates may be artificially
high. At various times since September 11, 2001, all of the services have
employed stop loss policies for both active duty personnel and reservists
at different times and in different ways. Stop loss policies include, but
are not limited to, the following examples. The Air Force was the first
service to use "stop loss" and invoked it for the entire Air Force between
September 2001 and January 2002. The Navy soon followed by implementing
stop loss for certain specialties between September 2001 and February
2002. The Marine Corps instituted stop loss for certain specialties
beginning in December 20, 2001, and terminated all stop loss in May 2003.
The Army began using stop loss in November 2001 and, as of the issuance of
this report, the Army is the only service operating a stop loss policy.24
Many within and outside of DOD are concerned about the impact that lifting
stop loss in the Army will have on attrition. While some fear that
servicemembers will leave the military in great numbers, others encourage
caution, suggesting that while a spike in attrition might occur, it would
be prudent to wait to see if the spike is temporary or long lasting. It is
important to note that a number of Army personnel affected by the stop
loss policy had already planned to leave the service upon completion of
their military obligation. Thus, some service members' departure is simply
delayed.

DOD should not only be able to measure the degree to which its retirement
system is meeting its goals but also assess how well the retirement system
helps the organization achieve its goals and pursue its

23 Until 1984, the amount appearing in the DOD budget under military
retirement reflected the annual amount paid to current retired military
personnel or their survivors. In 1984, Congress directed DOD to switch to
an accrual method for accounting for military retirement in the budget
process. (10 U.S.C. 1465(c).) An accrual method attempts, instead, to
reflect the liability arising from future retirees in current budgets.
This mechanism assumes that DOD will contribute a constant percentage of
an individual servicemember's pay annually to the accrual account.

24 Current Army policy prevents active and reserve component members from
leaving the service from the time their unit is alerted until 90 days
after their unit returns from active duty.

mission. This assessment would then enable DOD to determine if changes to
its reserve retirement system are needed from a workforce management
perspective. The fact that DOD has established attrition goals for its
reserve enlisted personnel is helpful. However, establishing goals for the
subset of attrition that is attributable to retirement would provide DOD
with even more meaningful data. For example, if overall attrition remained
constant but retirement attrition decreased, this trend could suggest that
the attrition was occurring at earlier career points and that fewer
servicemembers were deciding to make the military reserves a career. We
have previously reported that high-performing organizations identify their
current and future human capital needs and then create strategies for
identifying and filling gaps.25 The types of data that can inform
workforce planning efforts include, but are not limited to, the size and
shape of the workforce, skills inventory, attrition rates, and projected
retirement rates and eligibility. We previously testified26 that it is
important to consider the impact of retirement on the age and experience
distribution of the force, but that DOD lacked critical data needed to
assess alternatives to the existing retirement system. DOD also testified
that retirement benefits help shape the force by creating a reserve force
with certain characteristics and a specific distribution of personnel by
rank and years of service. However, officials in DOD's Office of the
Assistant Secretary of Defense for Reserve Affairs told us that they
currently do not have any way to empirically measure the retention effects
of the reserve retirement system. According to a 2001 study conducted for
the 9th Quadrennial Review of Military Compensation, DOD should (1) assess
whether the current skill, experience, and age composition of the reserves
are desirable and, if not, what they should look like now and in the
future and (2) develop an accession and retention model to evaluate how
successful varying combinations of compensation and personnel management
reforms would be in moving the reserves toward that preferred composition.

25 See GAO-02-373SP. 26 See GAO-03-549T.

Analysis of DOD Workforce Data Does Not Suggest That Changes to the
Reserve Retirement System Are Warranted

Our analysis of DOD workforce profile data over three points in time-
1991, 1997, and 200327-for enlisted and officer reservists, by years of
service, age, and pay grade, does not demonstrate that significant gaps
have occurred in the reserve component workforce that might suggest that
changes to the reserve component retirement system are warranted from a
workforce management perspective. In fact, our analysis revealed that, on
average, reserve enlisted personnel had more years of service and were
slightly older in 2003 than in 1991. For example, the average years of
service of reserve enlisted personnel was 1 to 3 years higher, for all
components, in calendar year 1997 than it was in calendar year 1991 and,
with the exception of the Marine Corps Reserve (which dropped by 1 year),
the average years of service was the same at the end of calendar year 2003
as it was at the end of calendar year 1997. Correspondingly, the average
age for all enlisted reservists was 1 to 3 years older in calendar year
1997 than in calendar year 1991. For the most part, average pay grades
remained constant over the three points in time. Similarly, our analysis
of reserve officers revealed that, on average, they also had more years of
service and were slightly older in 2003 than in 1991. Their average pay
grades remained constant across all components and points in time. More
detailed force profile data are presented in appendix III.

Reservists Compared to Active Duty Counterparts

In general, reservists had more years of service and were older than their
active duty counterparts. In calendar year 2003, for example, reserve
officer and enlisted personnel served an average of 17 and 11 years,
respectively, compared to active duty officers and enlisted personnel, who
served an average of 11 and 8 years, respectively. The average age of
reserve officers was 40 in calendar year 2003, and the average age of
active duty officers was 35. The average age of reserve enlisted personnel
was 33 in calendar year 2003, and the average age of active duty enlisted
personnel was 28. The average pay grade of reserve officers was O-3 in
calendar year 1991 and O-4 in 1997 and 2003, while the average pay grade
for active duty officers was O-3 in all three calendar years. The average
reserve enlisted pay grade was E-5 in all three calendar years. The
average active duty enlisted pay grade was E-4 in calendar year 1991 and
E-5 in calendar years 1997 and 2003.

27 The three points in time-1991, 1997, and 2003-were selected because
they coincided with major military events-Operations Desert Shield/Desert
Storm, military operations in Southwest Asia and the Balkans, and current
military operations in Iraq and Afghanistan, respectively.

Retention Rates Have Remained Fairly Constant Overall

An analysis of retention rates also did not indicate problems to suggest
that changes to the reserve retirement system are needed. Retention of
reservists between fiscal year 1991 and fiscal year 2003 remained
relatively stable, fluctuating between 76.5 and 82.9 percent. Retention
rates for reserve enlisted personnel were just slightly lower, fluctuating
between 75.1 and 82.3 percent, and retention rates for officers were
somewhat higher, ranging between 83.7 and 87.4 percent. Figure 3 shows
reserve retention rates for officers, enlisted personnel, and all
reservists from fiscal year 1991 through fiscal year 2003.

Figure 3: Total DOD Reserve Component Retention Rates, and Retention Rates
for Officer and Enlisted Reservists, Fiscal Years 1991 through 2003

Percentage

90

88 87.4 86.8

86

84

82

80

78

76 74 75.1

72

70

68 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Years

Officers
Total DOD reserve component
Enlisted

Source: GAO analysis of DOD data.

In general, the reserve components met DOD-established thresholds for
enlisted personnel attrition in recent years. In fiscal year 2003, all
components met their goals, with the exception of the Army National Guard
and the Air National Guard, which missed their targeted thresholds by less
than 1 percent. In fiscal years 2001 and 2002, all components met their
goals, except for the Army National Guard, which missed its targeted
thresholds by approximately 2 percent each year. Finally, in fiscal year

2000, all components met their targeted thresholds except for the Army
Reserve, which missed its threshold by less than 1 percent.

We also examined the career points at which attrition occurred.
Our analysis of retention data shows that relatively little attrition
occurred among reservists with more than 14 but less than 20 years of
service-those reservists most likely to be affected by retirement
policies.
(See appendix IV for more information on attrition of reserve personnel,
by years of service, for fiscal years 1991, 1997, and 2003.)

Retention Problems
Have Surfaced in
Some High-Demand
Occupational Specialties

DOD has, nonetheless, identified certain high-demand occupational
specialties where retention rates have decreased. These cases do not
necessarily suggest that uniform retirement reform is required. However,
they do suggest that DOD may need to take targeted corrective action,
using any number of human capital management tools available to DOD, to
address those specialties that are experiencing significant attrition.
Personnel in the fields of aviation, special forces, security,
intelligence, psychological operations, and civil affairs, for example,
have experienced a pace of operation two to seven times higher than those
of the average reservist. Retention rates for military police in the Army
Reserve, for example, decreased from approximately 67 percent in fiscal
year 2000 to 49 percent in fiscal year 2003. Similarly, retention
decreased for Army National Guard members in aviation, dropping from 80
percent in fiscal year 2000 to approximately 30 percent in fiscal year
2002. Additionally, from fiscal year 2000 to 2002, the retention rates of
Marine Corps Reserve members in civil affairs were consistently lower than
the overall retention rates for the entire Marine Corps Reserve.

Recent DOD-wide Survey Responses by reservists to questions about career
intention in recent Responses Reflect surveys administered by DOD indicate
that, if given a choice, most Retention Data reservists intended to
continue their service in the reserves. These surveys

were conducted during 2000 and in May and September of 2003.28

Responses by reservists in the 2000 survey indicated that 71.8 percent of
respondents intended to stay in the reserves. This percentage remained

28 A more recent survey was conducted in April 2004, the results of which
should be available by fall 2004, according to officials in the Office of
the Assistant Secretary of Defense for Reserve Affairs.

  Proposed Changes to Retirement System Are Costly and Need to Be Made in
  Context of the Total Force

relatively stable at 71.4 percent in May 2003, but it dropped to 67.3
percent in September 2003. Responses from reservists in each of the
components, who had been or were deployed overseas, however, showed
greater variation. For example, in the most recent survey
available-September 2003-67.3 percent of the reservists who responded to
the survey indicated that they would continue in the reserves, and a
majority of reservists noted that reserve retirement was a reason for
joining the reserves. However, only 52.6 percent of respondents in the
Army National Guard and 53.5 percent of respondents in the Army Reserve,
who were or had been called to active duty, responded that they would stay
in the Guard or Reserves.29 In the September 2003 survey, more than
one-half of the respondents cited the opportunity to earn a retirement
pension as a factor that influenced their decision to join the Guard or
Reserves. About 68.8 percent of higher graded officers (O-4 to O-6) and
66.5 percent of higher graded enlisted personnel (E-5 to E-9) cited the
reserve retirement system as a reason to join the Reserves, compared with
50.5 percent of lower graded officers (O-1 to O-3) and 37.5 percent of
lower graded enlisted personnel (E-1 to E-4).30

The estimated cost of current legislative proposals to change the reserve
retirement system is significant and will have a number of associated
effects. First, the estimated cost of lowering the age of receipt, from
age 60 to as low as 53 with limitations-or, in some cases, lower-will be
significant, given that all reservists who qualify for retirement-and not
simply the minority of reservists who have deployed to recent contingency
operations-would be covered under this change. Based on DOD data, our cost
estimates of the five current legislative proposals are substantial,
indicating that the cost of reserve retirement will increase by a range of
almost $3 billion to over $20 billion, depending on the proposal, over the
next 10 years. The Congressional Budget Office had not estimated the cost

29 The following question was asked in the 2000, May 2003, and September
2003 surveys: "Suppose that you have to decide whether to continue to
participate in the National Guard/Reserve. Assuming you could stay, how
likely is it that you would choose to do so?" Respondents were asked to
select from one of the following five response categories: very likely,
likely, neither likely nor unlikely, unlikely, very unlikely.

30 The following question was asked in the September 2003 survey: "How
much did each of the following contribute to your decision to join the
National Guard/Reserve?" Of the several different items listed after this
question, we analyzed responses to the following: "Opportunity to earn
credit toward a military pension." Respondents were asked to select from
one of the following five response categories: not at all, little
influence, some influence, great influence, very great influence.

of the retirement benefit for all five legislative proposals at the time
of our review. When associated health care costs are factored in, the
total projected increase ranges from more than $3 billion to nearly $35
billion over the same period. Considering the fact that military basic pay
will continue to grow and health care costs will continue to rise, the
longer-term cost will continue to be significant. Second, lowering the age
of receipt will align the reserve retirement system more closely with the
active duty retirement system, but may not affect reservists who are
experiencing increased deployments and associated hardships. Many of these
reservists are not likely to be the beneficiaries of a reduction in the
age of receipt of annuity, given that only one in four reservists
currently stays in the reserves long enough to become eligible for
retirement. Third, retirement is only one human capital management tool
available to DOD's workforce managers to compensate servicemembers for
hardships experienced during their careers. DOD's workforce managers can
consider making targeted investments in specific human capital approaches.
In addition to changes to the retirement system that would have a broad
effect, DOD also has more targeted approaches available to it. For
example, DOD has several special pays and allowances it can use to
compensate servicemembers, such as hazardous duty pay and family
separation allowances. In addition, Congress has approved, but DOD has yet
to implement, a special allowance for servicemembers who experience
excessive deployments. Fourth, DOD has recently identified a need to
rebalance the reserve and active duty forces to reduce the stress on
certain high-demand occupations. Better use of these human capital
management tools can help DOD address the issues of equity and the
increased use of reservists in military operations. Finally, DOD has not
yet studied changes to the reserve retirement system in the context of the
total force, even though these changes could have unintended consequences
on DOD's active duty forces. For example, if the reserve retirement system
were changed to offer an immediate annuity after 20 years of creditable
service, some personnel who have made career decisions to remain on active
duty until retirement might be inclined to leave active duty and apply to
serve their remaining time in a reserve component.

Legislative Proposals Are Members of the 108th Congress introduced five
legislative proposals that Designed to Lower the Age would reduce the age
at which retired reservists can receive their of Receipt annuities from
age 60 to something lower. The rationale often given for

these proposals is that reserve benefits, such as retirement, should be
enhanced in light of the frequency and duration of deployments that many
reservists are experiencing. One of these proposals would allow reservists

who served 20 years to retire and receive their annuities, immediately, at
any age. A second proposal would allow reservists who served 20 years to
leave the service and receive their annuities beginning at age 55. A third
proposal would reduce the age of receipt from 60 to as low as age 53-on a
sliding scale-by 1 year for every 2 additional years served beyond 20
years of service. For example, reservists who served 22 years would be
eligible to receive their annuities at age 59, and reservists who served
24 years would be eligible to receive their annuities at age 58. Under
this proposal, reservists would have to serve 34 years to be eligible to
receive annuities at age 53. A fourth proposal follows the same model but
establishes 55 as the minimum age at which reservists could begin to
receive annuities. Finally, a fifth proposal would retain eligibility of
receipt at age 60 but allow reservists to apply for receipt of reduced
annuities at younger ages. For example, the annuities of reservists who
applied to receive their annuities at age 55 would be reduced by 12.5
percent.

Estimated Cost Increases Will Be Substantial and May Not Target Reservists
Most Affected

The estimated cost of lowering the age of receipt of retirement annuities
will be substantial and has long-term budgetary implications. Based on
data provided by DOD's Office of the Actuary, we estimate that lowering
the age would cost DOD an additional $2.88 billion to $20.45 billion in
appropriated dollars, depending on the proposal, over the next 10 years.31
Reserve retirees would also be eligible, under four of the legislative
proposals, for DOD-provided health care benefits for each year that the
age of receipt is lowered. Using estimates of projected numbers of
retirees and assumptions regarding the demographic characteristics of
military personnel and their families provided by DOD's Office of the
Actuary, and average health costs for personnel and their dependents
provided by DOD's Office of Health Affairs, we calculated that DOD would
incur additional health care costs ranging from $2.39 billion to $14.24
billion.

Table 2 presents estimates of increased appropriations that would need to
be paid into the Military Retirement Fund to finance the increased
retirement benefits. The increased costs of even the most modest reduction
in the age of receipt can be expected to be largenearly $3 billion
between fiscal year 2005 and 2014. More generous proposals, such as
allowing reservists who have fulfilled 20 satisfactory years of service to
begin receiving retirement annuities immediately, could

31 The Congressional Budget Office had not estimated the cost of the
retirement benefit for all legislative proposals at the time of our
review.

increase appropriations by over $20 billion during fiscal years 2005 and
2014. Table 2 also presents estimated associated increases in health care
appropriations.

Table 2: Estimated 10-Year Increase in Appropriations of Five Legislative
Proposals between Fiscal Year 2005 and 2014

                              Dollars in billions

                                                 Age 53    Age 55 Age 55 with 
               Increased Immediate                 on a          on a reduced 
                   costs   annuity Age 55 sliding scale sliding scale annuity 
         Retirement fund    $20.45 $7.60          $2.88           $2.88 $3.59 
             Health care     14.24   6.03          2.49              2.39 NAb 
                  Totala    $34.69 $13.63         $5.37           $5.27 $3.59 

Source: GAO analysis of DOD data.

aThese dollar figures include the effects of inflation. When the effects
of inflation are removed, the total dollar estimates would be
approximately 8 percent lower. The net present value, or the funds that
would need to be invested in fiscal year 2005 to fund the additional
appropriations through fiscal year 2014, ranges from $4.28 billion to
$28.1 billion.

bThis proposal precludes members from receiving retiree health benefits
until age 60.

It should be noted that the numbers reported in table 2 underestimate the
costs of any of these proposals because they do not include the past
service liability created by the proposals. In addition, these estimates
assume that retention among personnel not yet eligible for retirement
would remain unchanged from recent experience. Officials in DOD's Office
of the Actuary suggested that improvements in retirement benefits could
lead to increases in retention and, therefore, produce a greater number of
retirees and higher costs than projected in this report.

Importantly, these costs are even more substantial, given the fact that
all reservists who qualify for retirement-and not just the minority of
reservists who have been called to active duty for recent contingency
operations-would be covered under any change to the reserve component
retirement system. Since 1991 and Operations Desert Shield/Desert Storm,
about one-fourth of all reservists who have served in the reserves were
called to active duty. Data, as of March 2004, show that about 28 percent
of the Ready Reserve have been called to active duty-either domestically
or abroad, for both short and long durations-in support of current
military operations since September 11, 2001. While all the services have
been affected, certain components and specialties have been affected more
than others. Furthermore, DOD's actuary has calculated that about only one
person in four who enters the reserves-

whether directly from civilian life or from active service-will retire
from the reserves. While lowering the age of receipt would align the
reserve retirement system more closely with the active duty retirement
system, it might not, as a result, affect those reservists who are
experiencing increased deployment and associated hardships.

Retirement Is Only One Tool Available to DOD's Workforce Managers

Retirement is only one human capital management tool available to shape
the workforce, and it is not clear how influential it is on a
servicemember's decision to leave or stay. That decision is a complex,
highly individual decision. Many factors, such as the availability of
civilian jobs, family considerations, and satisfaction with military life,
can influence a servicemember's decision. We previously found that, when
given a choice, servicemembers have tended to prefer immediate
compensation to compensation that is deferred.32 The Congressional Budget
Office found that new recruits would place a value of 4 cents on every
dollar to be paid to them in 20 years.33 Consequently, compensation such
as bonuses may be more effective at increasing overall retention than
changes to the reserve retirement system.

DOD has several special pays and allowances to further compensate
servicemembers who are experiencing excessive deployments. For example,
servicemembers who are away from their permanent duty station for more
than 30 consecutive days receive a family separation allowance to help
offset the additional expenses that may be incurred by their dependents.
In addition, servicemembers deployed for less than 30 days may be eligible
to receive regular per diem, which ranges from $20 to $533 per day,
depending upon the location. Furthermore, Congress has approved, but DOD
has yet to implement, a high-deployment allowance designed to compensate
certain servicemembers for lengthy as well as frequent shorter
deployments. There are also a number of special pays and allowances
intended to compensate servicemembers for hardships associated with
deployments and incentive pays to induce members to acquire a particular
skill or volunteer for certain career fields experiencing staffing
shortages.

32 See GAO, Military Retirement: Proposed Changes Warrant Careful
Analysis, GAO/T-NSIAD-99-94 (Washington, D.C.: Feb. 25, 1999).

33 See Congressional Budget Office, Military Compensation: Balancing Cash
and Noncash Benefits (Washington, D.C.: Jan. 16, 2004).

Another way to address possible retention problems in high-demand
occupations is to alleviate the stress placed on reservists by
transferring functions between active duty and reserve components. DOD
recently published a report identifying the need to rebalance the
forces-reserve and active duty-to ease stress on units and individuals
serving in occupational specialties that are in high demand, such as
military police, civil affairs and intelligence.34 DOD has undertaken
studies to balance its military forces by (1) moving skills and functions
that are found almost exclusively or predominately in the reserve
components, such as civil affairs, into the active force and (2)
reassigning personnel within the active and reserve components out of
low-demand specialties, such as heavy artillery, into high-demand
capabilities, such as military police and civil affairs. It should be
noted, however, that these efforts are in the early planning stages and
may take several more years to complete.

DOD Has Not Studied Changes to Reserve Retirement from a Total Force
Perspective

DOD has not yet studied changes to the reserve retirement system from a
total force perspective, even though it relies on the total force-both
active duty and reserve component personnel and units-to conduct its
military operations. Today, DOD cannot conduct operations without
contributions from both active duty and reserve personnel, given that
missions no longer reside solely within the active duty forces. DOD has
transferred primary missions to the reserve components and, accordingly,
reservists no longer simply augment active duty forces in times of
conflict. DOD has identified a need to rebalance functions between the
active duty and reserve forces to meet its mission. Similarly, DOD must
also balance the needs of both segments of its uniformed personnel when it
considers changes to manage its force.

Changes to the reserve component retirement system will have a number of
ramifications, some more difficult than others to foresee and potentially
unintended. Most importantly, not all of these ramifications will appear
in the retention experience of the reservists. While reducing the age of
receipt in the reserve retirement system may more closely align this
system with the active duty retirement system, this change could adversely
affect active duty retention rates. For example, if the reserve retirement
system were changed to offer an immediate annuity after 20 years of

34 See DOD, Office of the Deputy Assistant Secretary of Defense for
Reserve Affairs for Readiness, Training, and Mobilization, Rebalancing
Forces: Easing the Stress on Guard and Reserve (Washington, D.C.: Jan. 15,
2004).

Conclusions

creditable service, personnel who have made career decisions to remain on
active duty until retirement might be inclined to leave active duty and
apply to serve their remaining time in a reserve component or transfer to
the Inactive Ready Reserve and obtain retirement eligibility by
participating in and completing voluntary training programs. Although DOD
has not studied changes to the reserve retirement system from a total
force perspective,35 it has contracted RAND to prepare a study, due in
September 2004, on how changes to the reserve retirement system might
affect reserve retention and force management.

DOD is not in the best position to make a business case for endorsing or
dissuading efforts to adopt any of the legislative proposals that are
under consideration because it has not established expectations and
associated metrics that would help determine if changes to the reserve
retirement system are needed. In the absence of identified weaknesses in
the reserve component profiles or overall retention rates, it is difficult
to determine if problems exist that would be best served by changing the
reserve retirement system. Furthermore, changes to the retirement system
are expected to be costly, and retirement-a one-size-fits-all compensation
tool-does not necessarily target those reservists who are experiencing
frequent and lengthy deployments and associated hardships. This is
especially significant, given the fact that less than one-third of all
reservists have been called to active duty in recent years and that only
one in four reservists currently remains in the reserves long enough to
become retirement eligible. In other words, changes to the reserve
retirement system, at a substantial cost, would be extended to many
reservists who have not deployed and might only reach a minority of those
reservists who have deployed. Finally, DOD has not studied the impact of
proposed changes to the reserve component retirement system on the active
duty forces. DOD has transferred primary missions to the reserve
components and, accordingly, reservists no longer simply augment active
duty forces in times of conflict. DOD has already identified a need to
rebalance functions between the active duty and reserve forces to meet its
mission. Similarly, DOD must also balance the needs of both segments of
its uniformed personnel when it considers changes to manage its force.

35 According to the Office of the Under Secretary of Defense for Personnel
Readiness, DOD initiated a contract to study retirement issues from a
total force perspective. However, this contract was terminated and the
study was not completed.

  Recommendations for Executive Action

Agency Comments

Proposals, such as those that have been introduced and that would
permanently expand benefits, should be considered in the context of the
serious fiscal challenges facing this country. The longer-term costs and
implications of legislative proposals must be given adequate consideration
before they are enacted into law. The deficit is now so large that the
United States will not be able to grow its way out of it. Difficult
choices are inevitable. Congress and DOD have already implemented a number
of changes to improve the pay and benefits for reservists over the past
few years. Furthermore, DOD is exploring other human capital management
options. If changes are still warranted, opportunities may exist to apply
other workforce management tools and target corrective actions to those
reservists who are in military occupational specialties experiencing the
highest attrition rates.

We recommend that the Secretary of Defense direct the Under Secretary of
Defense for Personnel and Readiness, in concert with the Assistant
Secretary of Defense for Reserve Affairs, to take the following four
actions:

1. 	specify desired metrics for measuring the retention or attrition rates
of senior officer and enlisted reserve component personnel who are
approaching retirement eligibility and, therefore, are most likely to be
affected by changes to DOD's reserve component retirement system;

2. 	determine if gaps exist between the desired and actual rates of
attrition;

3. 	identify changes, if any, to the current reserve component retirement
system that would address these gaps, to the extent that they exist; and

4. 	evaluate any changes to the reserve component retirement system and
their associated long-term costs in the context of the total force.

In commenting on a draft of this report, DOD concurred with each of our
four recommendations. DOD also cited its ongoing contract with RAND to
conduct a detailed study of the reserve retirement system. DOD stated that
the contract calls for RAND to develop dynamic retention models that DOD
will be able to apply to evaluate the effect of any changes to the reserve
retirement system. DOD added that RAND is near completion of this study.
While the work that RAND is doing may provide DOD with

tools it needs to evaluate the effects of changes to the reserve
retirement system, we continue to emphasize that these evaluations should
be conducted in the context of the total force--both reserve and active
duty personnel. DOD's comments are reprinted in their entirety in appendix
V.

We are sending copies of this report to interested congressional
committees, the Secretary of Defense, and the Director of the Office of
Management and Budget. We will 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 regarding this report, please
contact me at (202) 512-5559 or [email protected] or Brenda S. Farrell at
(202) 512-3604 or [email protected]. Key contributors to this report were
Joseph Applebaum, William J. Doherty, Mark B. Dowling, Ann M. Dubois,
Kenya R. Jones, Karen L. Kemper, Kristy L. Kennedy, Stephen P. Marrin,
William McNaught, David E. Moser, Charles W. Perdue, Jennifer R.
Popovic, Jay Willer, and Gregory H. Wilmoth.

Derek B. Stewart
Director, Defense Capabilities and Management

                       Appendix I: Scope and Methodology

To assess the Department of Defense's (DOD) reserve component retirement
system as compared with certain aspects of the active duty retirement
system, we reviewed the legislative history and purpose of DOD's
retirement systems-both active and reserve-and DOD reserve retirement
policies and studies conducted by DOD and others that addressed reserve
retirement issues. Specifically, we reviewed DOD's Military Compensation
Background Papers and the results of the Sixth Quadrennial Review of
Military Compensation, conducted in 1988, that specifically addressed
retirement issues. We also interviewed officials in the Office of the
Assistant Secretary of Defense for Reserve Affairs and officials in each
of DOD's six reserve components-the Army Reserve, the Army National Guard,
the Naval Reserve, the Air Force Reserve, the Air National Guard, and the
Marine Corps Reserve.

To assess the extent to which data available to DOD suggest that changes
to the reserve component retirement system might be warranted, we analyzed
recent trends in overall workforce profiles and retention. For example, we
collected and analyzed data contained in DOD's Reserve Components Common
Personnel Data System and DOD's Active Duty Military Personnel Master File
and developed overall workforce profiles by years of service, age, and pay
grade. We did this for both officer and enlisted personnel on active duty
in each of the services and for officer and enlisted personnel reservists
serving in each of DOD's six reserve components. We also used year-end
data for three points in time-1991, 1997, and 2003-because these points in
time represent key periods when reservists were called to active duty to
support military operations. These included Operations Desert
Shield/Desert Storm, military operations in the Balkans and Southwest
Asia, and current operations in Iraq and Afghanistan. We then compared
average years of service, age, and pay grade over the three points in time
to identify the extent to which DOD may have experienced declines in any
of the variables we analyzed. In addition, we analyzed data provided by
the Office of the Assistant Secretary of Defense for Reserve Affairs and
generated from the Reserve Components Common Personnel Data System to
identify the retention and attrition patterns of reservists from fiscal
year 1991 through fiscal 2003. These data also included information on
certain high-demand military occupational specialties that experienced
higher than normal attrition between fiscal year 1991 and 2003. DOD
calculated the attrition rates by dividing total losses in a given year by
the number of reservists present at the end of each year, and then
compared retention rates over the years that we selected. We developed
retention rates for reserve officers, reserve enlisted personnel, and for
reservists as a whole, and then compared rates over time. We also analyzed
attrition by the years of service in which the

Appendix I: Scope and Methodology

attrition occurred, for the three points in time, to determine if
attrition was occurring at different stages in reservists' careers. We
interviewed officials in the Office of the Assistant Secretary of Defense
for Reserve Affairs and in each of DOD's reserve components to discuss
issues surrounding workforce profiles and retention.

We also reviewed DOD attitudinal surveys that were administered to
reservists in calendar years 2000 and 2003 and analyzed responses to
questions that addressed the intent of reservists to stay in or leave the
reserves. Response rates for the 2000 Survey of Reserve Component
Personnel, the May 2003 Status of Forces Survey of Reserve Component
Members, and the September 2003 Status of Forces Survey of Reserve
Component Members were 47.3 percent, 35.6 percent, and 31.9 percent,
respectively. There is a potential for bias in the estimates to the extent
that respondents and nonrespondents had different opinions on the
questions asked. Each survey was a random probability sample, so each was
only one of a number of large samples that could have been drawn. All
percentage estimates cited from each of the surveys have sampling errors
of plus or less than 5.6 percentage points. We used the weighting factors
and the sampling error methodology provided by the Defense Manpower Data
Center to develop estimates and sampling error estimates for each survey.

We assessed the accuracy and reliability of the Active Duty Military
Personnel Master File and the Reserve Components Common Personnel Data
Systems, from which all workforce profile, retention, and attrition data
were generated. Through written responses provided from the Defense
Manpower Data Center, which maintains these databases, we reviewed (1)
existing information about the data and the systems that produce them, (2)
the completeness of the electronic data, (3) DOD's methods of data
collection verification and quality control practices, (4) previous
quality reviews of the data and the frequency of these reviews, (5) any
limitations to the data, and (6) practices and controls over data entry
accuracy. We also interviewed agency officials knowledgeable about these
data systems. We determined that the data were sufficiently reliable for
purposes of this report. We also assessed the reliability of survey
responses for reporting purposes. The surveys we analyzed were based on
stratified, random samples of reserve component members, and the results
were weighted to achieve statistical validity. Again, the results of the
surveys were sufficiently reliable for purposes of this report.

To evaluate the potential effects and costs of legislative proposals that
address reserve retirement reform, we reviewed legislative proposals

Appendix I: Scope and Methodology

submitted in the 108th Congress and met with and obtained data on cost
projections-and the assumptions behind those projections-from the Office
of the DOD Actuary, the Office of the Assistant Secretary of Defense for
Health Affairs, and the Office of the Assistant Secretary of Defense for
Reserve Affairs. The Congressional Budget Office had not estimated the
cost of the retirement benefit for all five legislative proposals at the
time of our review. In addition, we reviewed reports developed by RAND and
others that analyze the potential effects of provisions in the legislative
proposals on the retention of reservists. We also acquired reports that
allowed us to assess alternative compensation and workforce management
methods for maintaining or increasing retention. Finally, we interviewed
officials in the Offices of the Assistant Secretary of Defense for Reserve
Affairs and Under Secretary of Defense for Personnel and Readiness to
analyze the extent to which DOD has studied reserve retirement reform from
a total force perspective.

We developed our own estimates of the increased costs that would be
incurred under each retirement proposal using projections made by DOD's
Office of the Actuary and data received from DOD's Office of Health
Affairs. We obtained estimates from the DOD actuary on the number of
personnel that would be affected by each retirement proposal and then
multiplied this population estimate by the estimated costs of providing
retirement and health care benefits.

To estimate the retirement benefit costs, we asked DOD's Office of the
Actuary to estimate the number of new persons who would be eligible to
retire under each retirement proposal. The DOD actuary also estimated the
normal cost of each proposal.1 To develop estimates of the normal cost of
each alternative, the DOD actuary projected future personnel levels based
upon the number of persons participating in the selected reserve

1 The normal cost of the military reserve retirement system is defined as
the percentage of reservists' basic salary that must be invested so that
sufficient resources will be available to finance all anticipated
retirement benefits, including costs of providing benefits to survivors.
The increased normal cost reflects the increase in the expected value of
the reserve retirement system expressed as a level percentage of
reservists' pay over their military careers. In other words, if this
percentage of reservists' pay were put aside and returned the rate of
return assumed by DOD's Office of the Actuary, it would be sufficient to
pay for the increased benefits. However, these normal costs will be paid
only for years 2005 and later. To have financed the increased benefits
contained in the proposals completely, it would have been necessary to
collect the increased normal costs over the previous years in which the
current group of reservists served. This was not done and, therefore, an
additional unfunded liability would have to be added to the Military
Retirement Fund.

Appendix I: Scope and Methodology

forces on September 30, 2003. The DOD actuary generally assumed that
future retention rates in the reserve forces would be the same rates that
were observed from 1997 to 2000 and also assumed that personnel, newly
eligible under the different proposals, would retire at the same rates as
personnel who are currently eligible. The DOD actuary assumed that
retirement benefit increases would vary from 1.3 to 3.0 percent between
fiscal year 2005 and 2014, that future pay raises for the uniformed
services would range from 3.4 to 3.75 percent, that the real rate of
return on retirement investments would be 3.25 percent, and that the rate
of discount would be 6.25 percent. For each proposal, we applied the DOD
actuary's estimate of the increase in the normal cost percentage to the
projected level of basic pay that would be paid to reservists from fiscal
year 2005 through fiscal year 2014 to estimate the amount of additional
funds that would need to be appropriated to the Military Retirement Fund.

To estimate the increased costs of providing health care benefits to newly
eligible retirees under each proposal, we again used the DOD actuary's
estimate of the number of new retirees under each proposal. We also used
assumptions, provided by DOD's Office of Health Affairs, that one-half of
these retirees would be single and that one-half would have dependents.
DOD's Office of Health Affairs assumed that the average number of
dependents in this latter group of retirees would be 2.94. The Office of
Health Affairs also assumed that the costs to provide health care would be
$2,242 for each retiree and $1,910 for each dependent in fiscal year 2005.
Based on these data, we calculated that the costs of providing health care
through DOD's health care system would increase 10 percent annually from
fiscal year 2006 through fiscal year 2014.

We conducted our review from November 2003 through July 2004 in accordance
with generally accepted government auditing standards.

Appendix II: Key Provisions of the Reserve Retirement Points System

The reserve retirement system, which has remained largely unchanged since
its enactment in 1948, is built upon a points system that provides DOD
with a means to convert the total number of years served by reservists on
a part-time basis into the equivalent number of full-time years served.
Essentially, DOD divides the total number of points that a reservist
earned in each qualifying year by 360-roughly, the number of days in a
year. This prorated adjustment provides DOD with the ability to calculate
retirement annuities for both active duty and reserve retirees on a
comparable basis.

Reservists must accumulate 20 years of creditable service to become
retirement eligible and, as of October 1, 2002, reservists must also serve
the last 6 years in a reserve component.1 A creditable year of service is
a year in which a reservist earns a minimum of 50 points. Retirement
points can be classified as inactive duty points and active duty points.

Reservists typically earn inactive duty points in a number of different
ways:

o  	Weekend drills. Reservists attend weekend drills each month. Each day
is divided into two drill periods, of 4 hours in duration, and reservists
earn 1 point for each drill. This equates to 4 points for each weekend, or
48 points annually.

o  	Annual training. Reservists participate in 2 weeks of training
annually and typically earn 12 points for this training time.

o  	Affiliation with a reserve unit. Reservists earn 15 points annually
for being a member of a reserve unit.

o  	Additional points. Reservists can earn additional points for such
things as completion of accredited correspondence courses or participation
in

1 The number of final years that reservists must serve in a reserve
component is established in law, and the number has fluctuated between 6
and 8 years. Reservists who completed 20 years of creditable service prior
to October 5, 1994, were required to fulfill their last 8 years of service
in a reserve component. Reservists who completed 20 years of creditable
service between October 5, 1994, and December 31, 2001, were required to
fulfill their last 6 years of service in a reserve component. Reservists
who completed 20 years of creditable service between January 1, 2002, and
September 30, 2002, were required to fulfill their last 8 years of service
in a reserve component. From October 1, 2002, until the present time,
reservists are required to serve their last 6 years in a reserve unit.

Appendix II: Key Provisions of the Reserve Retirement Points System

funeral honors duty. Reservists, however, cannot apply more than 90
inactive duty points toward retirement in any given year.2

In addition, reservists receive 1 point for any day served on active duty
prior to separating from active duty and affiliating with a reserve
component. When called to active duty, reservists also earn active duty
points-1 point for each day served. Reservists can earn a maximum of 365
active duty points in any calendar year and a career maximum of 10,950
points or 30 years of creditable service. Reservists who are called to
active duty will accumulate more points and, therefore, will receive
larger retirement annuities, should they stay until retirement, than their
reserve counterparts who are not called to active duty.

2 Prior to October 30, 2000, reservists could only apply 75 inactive duty
points toward retirement per year.

                  Appendix III: Reserve Workforce Profile Data

We analyzed data contained in DOD's Reserve Components Common Personnel
Data System and developed overall workforce profiles- by years of service,
age, and pay grade-for each reserve component, and for officer and
enlisted personnel, to identify trends that might suggest that changes to
DOD's reserve retirement system are warranted from a workforce management
perspective. We selected three points in time- calendar years 1991, 1997,
and 2003-because they represent key periods when reservists were called to
active duty. These included Operations Desert Shield/Desert Storm,
military operations in the Balkans and in Southwest Asia, and current
operations in Iraq and Afghanistan. We did not find significant variations
in overall years of service, age, or pay grade that would suggest that
changes to DOD's reserve retirement system are currently needed from a
workforce management perspective.

Reserve Enlisted Personnel Average Years of Service

The average years of service of reserve enlisted personnel was 1 to 3
years higher, for all components, in calendar year 1997 than it was in
calendar year 1991. With the exception of the Marine Corps Reserve, the
average years of service was the same at the end of calendar year 2003 as
it was at the end of calendar year 1997. In the Marine Corps Reserve, the
average years of service was 1 year lower at the end of calendar year 2003
than it was at the end of calendar year 1997, but the same as it was at
the end of calendar year 1991. Figure 4 shows the average years of service
by reserve enlisted personnel for all components for calendar years 1991,
1993, and 2003.

                  Appendix III: Reserve Workforce Profile Data

Figure 4: Reserve Enlisted Personnel Average Years of Service for Calendar
Years 1991, 1997, and 2003

                          Average years of service 16

                                     15 15

Similarly, reserve enlisted personnel were, on average, 1 to 3 years
older, across all components, in calendar year 1997 than they were in
calendar year 1991. Reserve enlisted personnel in four of the components
were, on average, the same age or 1 year older in calendar year 2003 as
they were in calendar year 1997. In the Army National Guard, enlisted
reservists were, on average, 1 year younger in calendar year 2003 than
they were in calendar year 1997, but 1 year older than they were in
calendar year 1991. In the Marine Corps Reserve, enlisted reservists were,
on average, 1 year younger in calendar year 2003 than they were in
calendar year 1997, and the same age as they were in calendar year 1991.
Figure 5 shows the

                                       14

                                       12

                                      10 8

                                       6

                                       4

                                       2

                             0 1991 1997 2003 Years

Air National Guard
Army National Guard
Army Reserve
Naval Reserve
Air Force Reserve
Marine Corps Reserve
Source: GAO analysis of DOD data.

Reserve Enlisted Personnel Average Age

                  Appendix III: Reserve Workforce Profile Data

average age of reserve enlisted personnel for all components for calendar
years 1991, 1997, and 2003.

 Figure 5: Reserve Enlisted Average Age for Calendar Years 1991, 1997, and 2003

                                 Average age 40

                               37 3736 36 3535 35

                                       35

                                       30

                                       25

                                       20

                                       15

                                      10 5

                             0 1991 1997 2003 Years

Air National Guard
Army National Guard
Army Reserve
Naval Reserve
Air Force Reserve
Marine Corps Reserve
Source: GAO analysis of DOD data.

Reserve Enlisted Figure 6 shows the average pay grade for reserve enlisted
personnel for all Personnel Average Pay components for calendar years
1991, 1997, and 2003. Average pay grades Grade remained constant in four
of the six components. The average enlisted pay

grade was one pay grade higher in the Air National Guard in calendar year
2003 than it was in calendar years 1991 and 1997. The average pay grade in
the Air Force Reserve was one pay grade higher in calendar years 1997 and
2003 than it was in calendar year 1991.

                  Appendix III: Reserve Workforce Profile Data

Figure 6: Reserve Enlisted Personnel Average Pay Grade for Calendar Years
1991, 1997, and 2003

Average enlisted pay grade

7

66 6

6

5

4

3

2

1

0 1991 1997 2003 Years

Air National Guard
Army National Guard
Army Reserve
Naval Reserve
Air Force Reserve
Marine Corps Reserve
Source: GAO analysis of DOD data.

                  Appendix III: Reserve Workforce Profile Data

Reserve Officer Average The average years of service by reserve officers
in the Air Force Reserve

Years of Service 	remained constant in calendar years 1991, 1997, and
2003. In the remaining five components, the average years of service was 1
to 3 years greater in 2003 than it was in calendar year 1991. None of the
components experienced a decline in the average number of years served for
the three points of time that we analyzed. Figure 7 shows the average
years of service by reserve officer personnel for all components for
calendar years 1991, 1997, and 2003.

Figure 7: Reserve Officer Average Years of Service for Calendar Years
1991, 1997, and 2003

Average years of service

20

18

18

16

14

12

10

8

6

4

2

0 1991 1997 2003 Years

Air National Guard
Army National Guard
Army Reserve
Naval Reserve
Air Force Reserve
Marine Corps Reserve

                       Source: GAO analysis of DOD data.

                  Appendix III: Reserve Workforce Profile Data

Reserve Officer Average Age

Reserve officers were, on average, 1 to 2 years older in calendar year
1997 than they were in calendar year 1991, with the exception of the Air
Force Reserve where officers were, on average, the same age in both
calendar years. Similarly, reserve officers were, on average, 1 to 2 years
older in calendar year 2003 than they were in calendar year 1997, with the
exception of reserve officers in the Air National Guard, who were, on
average, the same age in both of those calendar years. Figure 8 shows the
average age of reserve officer personnel for all components for calendar
years 1991, 1997, and 2003.

Figure 8: Reserve Officer Average Age for Calendar Years 1991, 1997, and
2003 Average age 44

42

42

40

38

36

34

32

30 1991 1997 2003 Years

Air National Guard
Army National Guard
Army Reserve
Naval Reserve
Air Force Reserve
Marine Corps Reserve
Source: GAO analysis of DOD data.

                  Appendix III: Reserve Workforce Profile Data

Reserve Officer Average Figure 9 shows the average pay grade of reserve
officers for all

Pay Grade 	components for calendar years 1991, 1997, and 2003. Average pay
grades remained constant within each component.

Figure 9: Reserve Officer Average Pay Grade for Calendar Years 1991, 1997,
and 2003

Average officer pay grade

5

4 4444 4 4444 4 4444

4

3

2

1

0 1991 1997 2003 Years

Air National Guard
Army National Guard
Army Reserve
Naval Reserve
Air Force Reserve
Marine Corps Reserve
Source: GAO analysis of DOD data.

                  Appendix IV: Attrition Rates for Reservists

We analyzed data contained in DOD's Reserve Components Common Personnel
Data System, for reserve officers, enlisted personnel, and all reservists,
to determine the career points at which attrition occurs and the extent to
which the current reserve retirement system influences the decisions of
reservists to stay in or leave reserve duty. Again, we selected three
points in time-1991, 1997, and 2003-because they represent key periods
when reservists were called to active duty to support military operations.
These included Operations Desert Shield/Desert Storm, military operations
in the Balkans and Southwest Asia, and current operations in Iraq and
Afghanistan. In each of the years we selected, we found little attrition
among those reservists with more than 14 years of service but less than 20
years of service-those servicemembers most likely to be affected by
changes to the reserve retirement system. Figures 10, 11, and 12 show
reserve attrition by years of service for fiscal years 1991, 1997, and
2003, respectively.

                  Appendix IV: Attrition Rates for Reservists

6,000

5,500

5,000

4,500

4,000

3,500

3,000

2,500

2,000

1,500

1,000

0 1 2 3 4 5 6 7 8 9 1011121314151617181920a Years of service

Officer

Enlisted

Total DOD reserve component

Source: GAO analysis of DOD data. aIncludes reservists leaving with 20 to
30 years of service.

                  Appendix IV: Attrition Rates for Reservists

Number of reservists

10,000

9,500

9,000

8,500

8,000

7,500

7,000

6,500

6,000

5,500

5,000

4,500

4,000

3,500

3,000

2,500

2,000

1,500

1,000

500

0 0 1 2 3 4 5 6 7 8 9 1011121314151617181920a Years of service

Officer
Enlisted
Total DOD reserve component

Source: GAO analysis of DOD data. aIncludes reservists leaving with 20 to
30 years of service.

                  Appendix IV: Attrition Rates for Reservists

Number of reservists

10,000

9,500

9,000

8,500

8,000

7,500

7,000

6,500

6,000

5,500

5,000

4,500

4,000

3,500

3,000

2,500

2,000

1,500

1,000

500

0 1 2 3 4 5 6 7 8 9 1011121314151617181920a

Years of service

Officer

Enlisted

Total DOD reserve component Source: GAO analysis of DOD data. aIncludes
reservists leaving with 20 to 30 years of service.

              Appendix V: Comments from the Department of Defense

Note: Page numbers in the draft report may differ from those in this
report.

              Appendix V: Comments from the Department of Defense

                       Appendix VI: Related GAO Products

Military Pay: Army Reserve Soldiers Mobilized to Active Duty Experienced
Significant Pay Problems. GAO-04-911. Washington, D.C.: August 20, 2004.

Military Personnel: DOD Has Not Implemented the High Deployment Allowance
That Could Compensate Servicemembers Deployed Frequently for Short
Periods. GAO-04-805. Washington, D.C.: June 25, 2004.

Military Personnel: Active Duty Compensation and Its Tax Treatment.
GAO-04-721R. Washington, D.C.: May 7, 2004.

Military Personnel: Observations Related to Reserve Compensation,
Selective Reenlistment Bonuses, and Mail Delivery to Deployed Troops.
GAO-04-582T. Washington, D.C.: March 24, 2004.

Budget Process: Long-Term Focus is Critical. GAO-04-585T. Washington,
D.C.: March 23, 2004.

Military Pay: Army National Guard Personnel Mobilized to Active Duty
Experienced Significant Pay Problems. GAO-04-89. Washington, D.C.:
November 13, 2003.

Military Personnel: DOD Needs More Data to Address Financial and Health
Care Issues Affecting Reservists. GAO-03-1004. Washington, D.C.: September
10, 2003.

Military Personnel: DOD Actions Needed to Improve the Efficiency of
Mobilizations for Reserve Forces. GAO-03-921. Washington, D.C.: August 21,
2003.

Military Personnel: DOD Needs to Assess Certain Factors in Determining
Whether Hazardous Duty Pay is Warranted for Duty in the Polar Regions.
GAO-03-554. Washington, D.C.: April 29, 2003.

Military Personnel: Preliminary Observations Related to Income, Benefits,
and Employer Support for Reservists During Mobilizations. GAO-03-573T.
Washington, D.C.: March 19, 2003.

Military Personnel: Oversight Process Needed to Help Maintain Momentum of
DOD's Strategic Human Capital Planning. GAO-03-237. Washington, D.C.:
December 5, 2002.

Appendix VI: Related GAO Products

Military Personnel: Active Duty Benefits Reflect Changing Demographics,
but Opportunities Exist to Improve. GAO-03-935. Washington, D.C.:
September 18, 2002.

Exposure Draft: A Model of Strategic Human Capital Management,
GAO-02-373SP. Washington, D.C.: March 15, 2002.

(350459)

  GAO's Mission

Obtaining Copies of GAO Reports and Testimony

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's 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, Contact:

Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470

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

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

                                 RECYCLED PAPER

           Presorted Standard Postage & Fees Paid GAO Permit No. GI00

United StatesEURGovernment Accountability OfficeEURWashington, D.C.
20548-0001EUR

Official BusinessEURPenalty for Private Use $300EUR

Address Service RequestedEUR
*** End of document. ***