Military Attrition: DOD Could Save Millions by Better Screening Enlisted
Personnel (Chapter Report, 01/06/97, GAO/NSIAD-97-39).

Pursuant to a congressional request, GAO reviewed the attrition rates of
first term, active-duty military personnel who are separated within the
first 6 months of their enlistments, focusing on: (1) how much the
services could save by achieving their goals for reducing 6-month
attrition; (2) the adequacy of the data that the of the Department of
Defense (DOD) uses to allow it to establish realistic goals for reducing
attrition; and (3) the principal reasons that enlistees are separated
from the services while they are still in training.

GAO found that: (1) all the services agree that reducing attrition is
desirable; (2) three services have attrition-reducing targets ranging
from 4 to 10 percent; (3) if the services reach their goals, they would
realize immediate short-term annual savings ranging from around $5
million to $12 million; (4) the services may not be able to realize
savings through reductions in their related training and recruiting
infrastructure for many years, but possible long-term savings could
range from more than $15 million to $39 million; (5) despite the fact
that the services have these goals, DOD, at present, lacks consistent
and complete information on the causes of attrition; (6) implementing
arbitrary attrition-reduction goals could result in a reduction in the
quality of recruits; (7) DOD's primary database for managing attrition
cannot be used to adequately determine the reasons that recruits
separate and to set appropriate targets for reducing attrition for two
reasons: (a) the services interpret and apply DOD's uniform set of
separation codes differently because DOD has not issued directives on
how to interpret them; and (b) current separation codes capture only the
official reason that an enlistee leaves the service; (8) thousands of
recruits are separated in the first 6 months because the services do not
adequately screen applicants for disqualifying medical conditions or for
preservice drug use; (9) one reason that this screening is inadequate is
that recruiters do not have sufficient incentives to ensure that their
recruits are qualified; (10) thousands of recruits also are separated
who fail to meet minimum performance criteria; and (11) recruits have
problems meeting performance standards because they are not physically
prepared for basic training and because they lack motivation.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  NSIAD-97-39
     TITLE:  Military Attrition: DOD Could Save Millions by Better 
             Screening Enlisted Personnel
      DATE:  01/06/97
   SUBJECT:  Military recruiting
             Military training
             Military cost control
             Attrition rates
             Enlisted personnel
             Management information systems
             Medical examinations
             Military discharges
             Drug abuse
             Cost effectiveness analysis
IDENTIFIER:  DOD Delayed Entry Program
             Armed Forces Qualifications Test
             Army Basic Training Program
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed 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.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to the Chairman and the Ranking Member, Subcommittee on
Personnel, Committee on Armed Services, U.S.  Senate

January 1997

MILITARY ATTRITION - DOD COULD
SAVE MILLIONS BY BETTER SCREENING
ENLISTED PERSONNEL

GAO/NSIAD-97-39

Military Attrition

(703122)


Abbreviations
=============================================================== ABBREV

  AFQT - Armed Forces Qualification Test
  ASVAB - Armed Services Vocational Aptitude Battery
  DEP - Delayed Entry Program
  DMDC - Defense Manpower Data Center
  DOD - Department of Defense
  GAO - General Accounting Office
  MEPCOM - Military Entrance Processing Command
  MEPS - Military Entrance Processing Station
  TRADOC - Training and Doctrine Command

Letter
=============================================================== LETTER


B-270643

January 6, 1997

The Honorable Dirk Kempthorne
Chairman
The Honorable Robert C.  Byrd
Ranking Member
Subcommittee on Personnel
Committee on Armed Services
United States Senate

This report responds to the request of the former Chairman and the
current Ranking Member that we review the attrition rates of
first-term, active-duty military personnel who are separated within
the first 6 months of their enlistments.  Specifically, we (1)
calculated how much the services could save by achieving their goals
for reducing 6-month attrition, (2) determined the adequacy of the
data that the Department of Defense uses to allow it to establish
realistic goals for reducing attrition, and (3) identified the
principal reasons that enlistees are separated from the services
while they are still in training. 

Unless you announce its contents earlier, we plan to make no further
distribution of this report until 3 days after its issue date.  At
that time, we will send copies to the Secretaries of Defense, the
Army, the Navy, and the Air Force and the Commandant of the Marine
Corps.  We will also make copies available to others upon request. 

Please contact me at (202) 512-5140 if you or your staff have any
questions concerning this report.  Other major contributors to this
report are listed in appendix II. 

Mark E.  Gebicke
Director, Military Operations and
 Capabilities Issues


EXECUTIVE SUMMARY
============================================================ Chapter 0


   PURPOSE
---------------------------------------------------------- Chapter 0:1

For at least the last decade, about one-third of enlistees in the
military services have failed to complete their first tours of duty. 
Concerned that the attrition rate was so high, the former Chairman
and the Ranking Member of the Subcommittee on Personnel, Senate
Committee on Armed Services, asked GAO to review the attrition rates
of first-term, active-duty military personnel who are separated
within the first 6 months of their enlistments.  Specifically, GAO
(1) calculated how much the services could save by achieving their
goals for reducing 6-month attrition, (2) determined the adequacy of
the Department of Defense's (DOD) data for allowing it to establish
realistic goals for reducing attrition, and (3) analyzed the
principal reasons that enlistees are separated from the services
while they are still in training. 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

After recruiters have prescreened applicants for the military
services, the applicants are sent to military entrance processing
stations (MEPS), which are the responsibility of the Military
Entrance Processing Command.  When it has been determined that the
applicants are qualified, through medical and aptitudinal tests, they
are sworn into the Individual Ready Reserve, in an unpaid status, for
up to 1 year.  Once they are called to active duty, enlisted
personnel enter basic training, which can last from 6 to 12 weeks,
depending on the service.  After basic training, recruits go on to
initial skill training, which can range from a few weeks to more than
a year. 

In fiscal year 1994, DOD recruited more than 176,000 new recruits. 
Of that number, more than 25,000 were separated by the 6-month point
in their contracts. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

All the services agree that reducing early attrition is desirable. 
To this end, three services have attrition-reducing targets ranging
from 4 to 10 percent.  If the services reach their goals, they would
realize immediate short-term annual savings ranging from around $5
million to $12 million.  The services may not be able to realize
savings through reductions in their related training and recruiting
infrastructure for many years.  However, possible long-term savings
could range from more than $15 million to $39 million.  Despite the
fact that the services have these goals, DOD, at present, lacks
consistent and complete information on the causes of attrition. 
Implementing arbitrary attrition-reduction goals could result in a
reduction in the quality of recruits. 

DOD's primary database for managing attrition cannot be used to
adequately determine the reasons that recruits separate and to set
appropriate targets for reducing attrition for two reasons:  (1) the
services interpret and apply DOD's uniform set of separation codes
differently because DOD has not issued directives on how to interpret
them and (2) current separation codes capture only the official
reason that an enlistee leaves the service. 

Thousands of recruits are separated in the first 6 months because the
services do not adequately screen applicants for disqualifying
medical conditions or for preservice drug use.  One reason that this
screening is inadequate is that recruiters do not have sufficient
incentives to ensure that their recruits are qualified.  Thousands of
recruits also are separated who fail to meet minimum performance
criteria.  Recruits have problems meeting performance standards
because they are not physically prepared for basic training and
because they lack motivation. 


   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4


      DOD COULD SAVE MILLIONS OF
      DOLLARS BY REDUCING
      ATTRITION
-------------------------------------------------------- Chapter 0:4.1

If the services reach their goals for reducing attrition, they would
realize immediate short-term savings because they would be
transporting, feeding, clothing, and paying fewer recruits.  In some
cases, reducing attrition may require that the services add
preenlistment medical tests or more screening mechanisms to their
recruiting and examining processes.  However, GAO believes that these
added costs would be more than offset by the immediate short-term
savings achieved through having to recruit, process, and train fewer
recruits.  Even larger dollar savings could be realized over time as
the services began to reduce the infrastructure associated with
recruiting and training enlistees. 

Using GAO's calculations of the fixed and marginal costs of
recruiting and training and the services' highest and lowest targets
for reducing attrition, GAO estimates that if the services were to
reduce their 6-month attrition by 4 percent, their immediate
short-term savings would be $4.8 million.  If the services achieved a
10-percent reduction of attrition, their short-term savings would be
$12 million.  Over time, if the services reduced 6-month attrition by
4 percent, their infrastructure savings could be as high as $15.6
million.  If they were able to reduce their 6-month attrition by
10 percent, potential infrastructure savings could be as much as $39
million. 


      DOD'S DATA DOES NOT ALLOW
      THE SERVICES TO SET
      REALISTIC ATTRITION GOALS
-------------------------------------------------------- Chapter 0:4.2

While significant savings could be achieved by reducing attrition,
GAO believes that the services' current goals for reducing attrition
are arbitrary.  That is, DOD and the services do not currently have
sufficient information to determine what portion of 6-month attrition
is truly avoidable.  The danger of setting arbitrary goals is that
these goals can become "attrition ceilings," which can result in the
inadvertent retention of lower quality recruits.  To set realistic
and achievable targets for reducing attrition, DOD and the services
need more complete and accurate data on why recruits are being
separated. 

DOD's current data on attrition is inconsistent and incomplete for
two reasons.  First, the services interpret DOD's definitions of
separation codes differently and therefore place enlistees with
identical situations in different discharge categories.  Second,
DOD's separation codes capture only the officially assigned reason
for discharge, when many other factors may result in an enlistee's
separation.  DOD has not issued guidance for applying these
separation codes. 


      SCREENING PROCESSES DO NOT
      IDENTIFY THOUSANDS OF
      RECRUITS WHO ARE UNQUALIFIED
      FOR SERVICE
-------------------------------------------------------- Chapter 0:4.3

About 83 percent of the 25,000 who were discharged in their first 6
months were assigned separation codes indicating that they (1) were
medically unqualified for military service, (2) had character or
behavior disorders, (3) had fraudulently or erroneously entered the
military, or (4) failed to meet minimum performance criteria.\1
Separations for medical conditions and failure to meet performance
standards represent at least 55 percent of all 6-month attrition for
enlistees who entered the services in fiscal year 1994.  This
percentage is understated for two reasons.  First, some persons who
have medical problems are categorized as fraudulent enlistments
because they concealed medical problems.  Second, some persons who
have performance problems are categorized as having character or
behavior disorders.  GAO was not able to calculate the number of
persons discharged for drug use because these separations are
categorized in many different ways. 

GAO found that recruits were enlisted and later separated because
DOD's screening processes were inadequate in the following ways: 

  -- Recruiters do not have adequate incentives to ensure that their
     recruits are qualified.  The Navy recently began to subtract
     points from recruiters' quotas when their enlistees did not
     graduate from basic training.  It is too soon, however, to
     determine the effect of this change on attrition.  Over the
     years, the Marine Corps has allowed its recruiting units the
     flexibility to tie recruiters' incentive systems to enlistees'
     successful completion of basic training.  However, this policy
     has not been uniformly applied throughout the Marine Corps, and
     its incentive system, as those of the other services, does not
     appear to provide adequate incentives for recruiters to screen
     out unqualified applicants.  Basic training personnel suggested
     that awarding recruiters with partial credit for screening out
     unqualified personnel or changing monthly goals to floating
     3-month goals might relieve the pressure on recruiters to enlist
     personnel later found to be unqualified. 

  -- The services do not require all applicants to provide the names
     of their medical insurers or their past medical providers. 
     Also, the medical screening forms contain vague and ambiguous
     questions and may be easy for applicants to falsify. 

  -- DOD's current system of capturing information on medical
     diagnoses does not allow it to track the success of recruits who
     receive medical waivers.  DOD has just approved a project to
     compile a comprehensive database of medical conditions for all
     accessions.  Information from this database will provide DOD
     with the ability to reevaluate its physical enlistment
     standards, to analyze the medical reasons that recruits are
     separated, to make fact-based policy changes to reduce medical
     attrition, and to determine whether it would be cost-effective
     to provide more medical tests to all or selected groups of
     applicants. 

  -- The responsibility for reviewing medical separation cases to
     determine whether medical conditions should have been detected
     at the MEPS now resides with the Military Entrance Processing
     Command, the organization responsible for the medical
     examinations. 

  -- The Navy and the Marine Corps do not test applicants for drugs
     at the MEPS but wait until they arrive at basic training. 


--------------------
\1 The Defense Manpower Data Center maintains data on all the
services' enlistees; fiscal year 1994 was the most current year for
which complete data was available. 


      THOUSANDS OF RECRUITS ARE
      DISCHARGED FOR FAILURE TO
      MEET MINIMUM PERFORMANCE
      CRITERIA
-------------------------------------------------------- Chapter 0:4.4

More than 7,200 of the recruits who entered the services in fiscal
year 1994 were discharged in the first 6 months of service because
they failed to meet minimum performance criteria.  Basic training
personnel throughout the services said that these recruits are not
physically prepared for basic training and lack motivation.  Basic
training personnel suggested that recruits might be better prepared
for the physical demands of basic training if they were more fully
informed of the services' physical training requirements and
encouraged to exercise to become physically fit before going to basic
training.  The Army has a new program, which was nearing
implementation in December 1996, to (1) award enlistees retirement
points for participating in physical activities with their recruiters
before going to basic training and (2) allow enlistees access to
military fitness centers and military medical facilities if they are
injured. 

To try to improve recruits' motivation, all the services have taken
actions to improve the basic training environment.  They have
established motivational and rehabilitation units for recruits with
motivational problems and injuries.  The Army and the Air Force, in
particular, have stressed positive leadership by their drill
instructors.  Despite these improvements, GAO's interviews with
separating recruits suggest that negative leadership techniques
continue to be a factor in recruits' lack of motivation to meet
performance standards. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:5

To reduce the attrition of enlisted personnel during the first 6
months of their terms of enlistment, GAO recommends that the
Secretary of Defense issue implementing guidance on DOD's separation
codes and direct the services to strengthen their recruiter incentive
and medical screening systems.  GAO also recommends that DOD use its
newly proposed database of medical diagnostic codes to improve
medical screening and that DOD move the responsibility for reviewing
medical separations from the Military Entrance Processing Command. 
Finally, GAO recommends that drug testing for all services be moved
to the MEPS and that the services adopt Delayed Entry Programs
similar to the Army's new proposed program.  These recommendations
are presented in their entirety in chapters 2 and 3. 


   AGENCY COMMENTS AND GAO'S
   EVALUATION
---------------------------------------------------------- Chapter 0:6

In commenting on a draft of the GAO report, DOD concurred with GAO's
recommendation to use DOD's newly proposed database on medical
diagnostic codes to improve medical screening and with GAO's
recommendation to strengthen the services' Delayed Entry Programs. 
DOD partially concurred with GAO's recommendations to issue
implementing guidance on DOD's separation codes and to direct the
services to strengthen their recruiter incentive systems and
screening mechanisms.  DOD also partially concurred with the GAO
recommendation to test all applicants for military service for drugs
before they report to basic training.  DOD did not concur with GAO's
recommendation to remove the review of medical separation files from
the agency that conducts the medical screening.  DOD believes that
the Military Entrance Processing Command is the appropriate entity to
perform this review.  GAO continues to believe that an entity
completely outside the medical screening process would be more able
to objectively determine whether the MEPS physicians should have
discovered disqualifying medical problems.  DOD's comments appear in
their entirety in appendix I and are discussed in
chapters 2 and 3. 


INTRODUCTION
============================================================ Chapter 1

For at least the last decade, about one-third of those who have
enlisted in the military services have failed to complete their
initial enlistment contracts.  One-third of these separating
enlistees left the military before they reported to their first duty
assignments.  The military services make a substantial investment in
training, time, equipment, and related expenses for military
enlistees.\1 The separation of enlisted personnel before they
complete their initial training is wasteful because the services lose
their investment and must increase accessions to replace these
losses.  Consequently, first-term attrition is an issue of
significant concern at all levels within the armed forces. 


--------------------
\1 Not all recruits have completed training at the 6-month point in
their first terms because some initial skill training lasts beyond
this point.  In rare cases, initial skill training can last as long
as a year or more.  However, for the purpose of this report, we
examined attrition at the 6-month point because at that time, most
enlistees have completed both basic and follow-on training and are
being assigned to their first duty stations. 


   A SIGNIFICANT PORTION OF
   FIRST-TERM ATTRITION OCCURS
   DURING TRAINING
---------------------------------------------------------- Chapter 1:1

New recruits take an enlistment oath and sign a contract to serve one
of the military services for a specified period of time, typically 4
years.  Despite this contractual obligation, Department of Defense
(DOD) data shows that about one-third of new recruits fail to
complete their first terms.  This attrition figure has been
relatively constant over the past
10 years and has held true for each of the services.  Table 1.1 shows
attrition rates for enlistees who entered the services in fiscal
years 1986 through 1991.  DOD generally tracks enlisted attrition up
to the 3-year point in enlistees' first terms.  In this report,
however, we show attrition at the 4-year point because the majority
of enlistees have 4-year contracts.  Calculations of attrition at the
3-year point do not include the attrition of those who have 4-year
contracts and leave the services in the last year of their
commitments.  Enlistees who entered the services in fiscal year 1986
were scheduled to complete 4-year contracts in fiscal year 1990. 
Likewise, enlistees who entered the services in fiscal year 1991 were
expected to complete their 4-year contracts in fiscal year 1995. 



                               Table 1.1
                
                    Percentage of Enlistees Who Are
                   Separated During Their First Terms

Fiscal year enlistees                       Marine     Air         All
entered the services          Army    Navy   Corps   Force    services
--------------------------  ------  ------  ------  ------  ----------
1986                          32.1    35.2    38.1    27.2        32.6
1987                          31.9    33.4    35.3    26.3        31.7
1988                          34.6    33.8    32.5    26.3        32.8
1989                          36.3    35.5    34.5    31.2        35.0
1990                          37.2    34.2    38.0    31.2        35.4
1991                          37.7    31.7    35.4    32.8        34.6
----------------------------------------------------------------------
Source:  Defense Manpower Data Center. 

Our analysis of the data further reveals that attrition is not evenly
distributed throughout a first-term enlistment.  About one-third of
first-term attrition occurs within the first 6 months of an
enlistee's term, during the time when many recruits are still in
training and before they report to their first duty assignments. 
Table 1.2 displays the attrition rates at the 6-month point, again
for each service and for all services, for enlistees who entered the
services between fiscal years 1986 and 1994.  (Fiscal year 1994 was
the latest year for which the Defense Manpower Data Center (DMDC) had
complete data at the time of our review.)



                               Table 1.2
                
                    Percentage of Enlistees Who Are
                Separated in the First 6 Months of Their
                              First Terms

Fiscal year enlistees                       Marine     Air         All
entered the services          Army    Navy   Corps   Force    services
--------------------------  ------  ------  ------  ------  ----------
1986                          10.4    13.1    15.9    10.7        11.8
1987                           9.2    12.7    13.2    10.0        10.8
1988                           9.8    14.4    12.6     9.0        11.6
1989                          10.0    12.8    13.9     9.4        11.3
1990                          10.7    10.1    15.6    10.2        11.1
1991                          13.0    10.2    14.1    10.5        11.9
1992                          12.8    12.9    12.9     9.2        12.3
1993                          15.3    15.8    13.6    11.6        14.6
1994                          15.7    15.7    12.5    11.6        14.4
----------------------------------------------------------------------
Source:  DMDC. 

Figure 1.1 shows that, in fiscal years 1990 through 1994, DOD's
2-month, 6-month, and 12-month attrition rates increased steadily in
a parallel pattern.  Attrition rates shown in figure 1.1 are
cumulative.  That is, 6- and 12-month attrition rates include all
attrition up to those two points in time. 

   Figure 1.1:  DOD's 2-Month,
   6-Month, and 1-Year Attrition
   for Fiscal Years 1990 Through
   1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of DMDC data. 


   DEMOGRAPHICS OF FISCAL YEAR
   1994 RECRUITS
---------------------------------------------------------- Chapter 1:2

We concentrated our analysis of 6-month attrition on the cohort of
recruits who joined the military in fiscal year 1994, as this was the
latest year for which DMDC had complete statistics during the time of
our review.  Additional data provided to us by DOD demonstrates that
fiscal year 1994 was a representative year in terms of the education
levels and quality of the recruits who joined the military. 
Researchers have investigated several factors that influence
attrition during the first term of enlistment.  These include
educational credentials, gender, age, race, enlistment term, and
military occupational specialties.\2 According to DOD and the
services, the most important of these variables in determining the
attrition rate is recruits' educational attainment.  Most researchers
have found that enlistees who were high school graduates had lower
attrition rates.  A second predictor of lower attrition rates is
enlistees' scores on the Armed Forces Qualification Test (AFQT). 
Those who score in the upper
50th percentile have historically had lower attrition rates.  In
fiscal year 1994, 96 percent of all recruits were high school
graduates, and 72 percent of all recruits scored in the upper 50th
percentile on the AFQT.  In that same year, 68 percent of all
recruits were high school graduates and scored in the upper 50th
percentile of the AFQT.  All of these figures compare favorably with
data from other recent fiscal years. 

The magnitude of DOD's 6-month attrition becomes apparent when
studied in context with DOD's total accessions.  In fiscal year 1994,
DOD recruited more than 176,000 recruits who did not have prior
military service.  Of that number, more than 25,000 recruits were
separated by the 6-month point in their contracts. 


--------------------
\2 See Report to Congress:  Educational Enlistment Standards: 
Recruiting Equity for GED Certificates (Office of the Assistant
Secretary of Defense (Force Management Policy), Apr.  1996);
Attrition Revisited:  Identifying the Problem and Its Solutions
(Human Resources Research Organization, FR-PRD-95-01, Jan.  1995);
Who Stays, Who Leaves?  Attrition Among First-Term Enlistees (Rand,
N-2967-FMP, May 1989); Trends in Attrition of High-Quality Military
Recruits (Rand, R-3539-FMP, Aug.  1988); and First-Term Attrition in
the Marine Corps (Center for Naval Analyses, CRM 92-200, Mar.  1993). 


   THREE SEPARATE COMMANDS
   RECRUIT, SCREEN, AND TRAIN NEW
   ENLISTEES
---------------------------------------------------------- Chapter 1:3

After a recruiter prescreens an applicant for military service, the
applicant is sent to one of 65 military entrance processing stations
(MEPS) located throughout the country.  At the MEPS, which are under
the direction of the Military Entrance Processing Command (MEPCOM),
the applicant takes the Armed Services Vocational Aptitude Battery
(ASVAB) to determine whether he or she is qualified for enlistment
and a military job specialty,\3

and a medical examination is given to determine whether he or she
meets physical entrance standards.  After it has been determined that
an applicant is qualified, the applicant is sworn into the service
and enters the Delayed Entry Program (DEP).  When an applicant enters
the DEP, he or she becomes a member of the Individual Ready Reserve,
in an unpaid status, and awaits being called to active duty.  An
individual may remain in the DEP for up to 1 year.  Just before
reporting to the service basic training command, the new recruit
returns to the MEPS, undergoes a brief physical examination, and is
sworn into active duty. 

Basic training lasts from 6 to 12 weeks.  Most enlistees have
completed basic training before the 3-month point in their first
terms, though their graduation points may vary, depending on how long
their basic training lasts and on whether they have to be held back
to repeat some parts of basic training.  The Air Force basic training
program lasts 6 weeks and is given at one training site, located at
Lackland Air Force Base in San Antonio, Texas.  Navy recruits remain
in basic training for 9 weeks at one site, located at the Naval
Training Center in Great Lakes, Illinois.  The Marine Corps' basic
training curriculum is 11 weeks long for men and
12 weeks long for women, and recruits are trained in San Diego,
California, and Parris Island, South Carolina.  The Army has two
types of basic training sites:  sites that provide only basic combat
training and one-station unit training sites that provide both basic
combat training and follow-on initial skill training.  The Army's
basic combat training sites are located at Fort Knox, Kentucky; Fort
Sill, Oklahoma; Fort Leonard Wood, Missouri; and Fort Jackson, South
Carolina.  The Army's one-station unit training sites are located at
Fort Benning, Georgia; Fort Knox, Kentucky; Fort Sill, Oklahoma; Fort
McClellan, Alabama; and Fort Leonard Wood, Missouri.  Army basic
training lasts 8 weeks. 

By the 6-month point in their first terms, most enlistees have
completed follow-on training in technical skills, though the length
of such training can vary widely, from a few weeks to a year or more. 
In some cases, graduates of basic training go directly to their first
duty assignments.  Figure 1.2 displays the most common recruiting and
training pipeline for new enlistees. 

   Figure 1.2:  Process of
   Recruiting and Training
   Enlisted Personnel

   (See figure in printed
   edition.)

Source:  GAO. 


--------------------
\3 In some cases, applicants are given the ASVAB in high school or at
independent sites apart from the MEPS. 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:4

At the request of the former Chairman and Ranking Member of the
Subcommittee on Personnel, Senate Committee on Armed Services, we
reviewed the attrition rates of first-term, active-duty military
enlistees who are separated from the military within the first 6
months of their enlistments.  Specifically, we (1) calculated how
much the services could save by achieving their goals for reducing
6-month attrition, (2) determined the adequacy of DOD's data for
allowing it to establish realistic goals for reducing attrition, and
(3) analyzed the principal reasons that enlistees are being separated
from the services while they are still in training. 

We limited the scope of our review to attrition at the 6-month point
for two reasons.  First, an enlistee's discharge is categorized as an
entry-level separation until the 6-month point in the enlistee's
term.  The entry-level, 180-day point serves in some sense as a
probationary period.  If enlistees are discharged after 6 months,
they may be entitled to more benefits and undergo a more complex
separation process.  Our second reason for measuring attrition at the
6-month point is that this point marks the end of training for most
first-term enlistees.  Because of the variation in the length of
follow-on training, however, some enlistees continue in training for
a year or longer into their first terms. 

To identify the potential cost savings that DOD could realize by
reducing its first-term attrition, we first determined the magnitude
of annual service accessions and first-term attrition, over time, by
obtaining and reviewing data maintained by DMDC.  We also compared
DOD- and service-provided data regarding average costs to recruit,
examine, test, screen, transport, and train new enlistees.  This data
includes both the short-term variable costs that are associated with
the cost per recruit and the longer-term fixed costs that are
associated with the infrastructure required to recruit and train new
enlistees.  In addition, we reviewed service-identified targets for
reducing first-term attrition and applied these targets to the cost
data to identify the potential for cost savings. 

To determine the adequacy of DOD's data regarding reasons for
first-term attrition, we analyzed DMDC's database of enlistee
separations and reviewed DOD's corresponding list of separation
codes, which designate the official reasons that enlistees are
separated.  Additionally, we reviewed the services' separation
instructions and met with personnel officials at basic training
locations for each of the services to identify similarities and
differences in the way the separation codes are applied at the
different locations. 

To analyze the principal reasons that DOD is separating enlistees
within the first 6 months of their enlistments, we reviewed DMDC's
database of separations in each of the services for enlistees who
entered the services in fiscal years 1990 through 1995.  We then
compared this data to the service separation codes.  Specifically, we
concentrated on separations that occurred in fiscal year 1994, as
this was the most recent year for which DMDC had complete data at the
time of our review.  To understand reasons for attrition, we also
interviewed officials in DOD and each of the services who are
involved in recruiting, examining, screening, training, and
separating enlistees.  To obtain the perspective of separating
recruits, we conducted one-on-one interviews with a total of 126
recruits, who were being separated but were still at the basic
training commands at the time of our site visits.  We recognize that
these recruits do not represent a statistical sample of all recruits
who will be separated this year.  Nevertheless, their responses do
supplement information provided to us by DOD and service officials. 

We performed our work at the following DOD and service headquarters,
commands, and installations: 

  -- Directorate for Accession Policy, Office of the Assistant
     Secretary of Defense, Force Management Policy, Washington, D.C.;

  -- Office of the Assistant Secretary of Defense, Health Affairs,
     Washington, D.C.;

  -- Army Directorate of Military Personnel Management, Washington,
     D.C.; U.S.  Army Recruiting Command, Ft.  Knox, Kentucky; and
     Army Basic Training, Fort Jackson, South Carolina;

  -- Air Force Directorate of Military Personnel Policy, Washington,
     D.C.; Air Force Recruiting Service, Randolph Air Force Base, San
     Antonio, Texas; and Air Force Basic Training, Lackland Air Force
     Base, San Antonio, Texas;

  -- Manpower Plans and Policy Division, Marine Corps Headquarters,
     Arlington, Virginia; Marine Corps Recruiting Command, Arlington,
     Virginia; and Marine Corps Basic Training, Marine Corps Recruit
     Depot, Parris Island, South Carolina;

  -- Navy Office of Military Personnel Policy and Career Progression,
     Washington, D.C.; Navy Recruiting Command, Arlington, Virginia;
     Navy Bureau of Medicine and Surgery, Washington, D.C.; and Navy
     Recruit Training Command, Naval Training Center, Great Lakes,
     Illinois; and

  -- U.S.  Military Entrance Processing Command, North Chicago,
     Illinois; Military Entrance Processing Station, Fort Jackson,
     South Carolina; and Military Entrance Processing Station,
     Milwaukee, Wisconsin. 

With regard to recruit training, we conducted audit work at Lackland
Air Force Base and the Great Lakes Naval Training Center because
those are the only locations where the Air Force and the Navy provide
basic training.  In the case of the Marine Corps, we selected Parris
Island because this is the only site where the Marines train both
male and female recruits.  In the case of the Army, we selected Fort
Jackson, South Carolina, because this training location provided the
greatest variation in job specialties.  We conducted our review from
November 1995 to October 1996 in accordance with generally accepted
government auditing standards.  DOD's comments on a draft of this
report are summarized in chapters 2 and 3 and are presented in their
entirety in appendix I. 


DOD COULD SAVE MILLIONS OF DOLLARS
BY REDUCING ATTRITION
============================================================ Chapter 2

All the services agree that reducing early attrition is desirable. 
To this end, three services have developed attrition-reduction
targets ranging from
4 to 10 percent.  If the services were to reach their goals, they
would realize immediate short-term savings because they would be
transporting, feeding, clothing, and paying fewer recruits.  In some
cases, reducing attrition may require that the services add
preenlistment medical tests or more screening mechanisms to their
recruiting and examining processes.  However, we believe that these
added costs would be more than offset by immediate short-term
savings.  The services could accrue these savings because they would
need to recruit, process, and train fewer recruits to meet the same
accession needs.  Even larger dollar savings could be realized over
time as the services began to reduce the infrastructure associated
with recruiting and training enlistees. 

Using our calculations of the fixed and marginal costs of recruiting
and training and the services' highest and lowest targets for
reducing attrition, we estimate that if the services were to reduce
their 6-month attrition by
4 percent, their immediate short-term savings would be $4.8 million. 
If the services achieved a 10-percent reduction of attrition, their
short-term savings would be $12 million.  Over time, if the services
reduced 6-month attrition by 4 percent, their infrastructure savings
could be as high as $15.6 million.  If they were able to reduce their
6-month attrition by
10 percent, potential infrastructure savings could be as much as $39
million. 

While we believe that significant savings could be achieved by
reducing attrition, we also believe that the services' current goals
for reducing attrition are arbitrary.  That is, DOD and the services
do not currently have sufficient information to determine what
portion of 6-month attrition is truly avoidable.  The danger of
setting arbitrary goals is that these goals can become "attrition
ceilings," which can result in the inadvertent retention of lower
quality recruits.  To set realistic and achievable targets for
reducing attrition, DOD and the services need more complete and
accurate data on why recruits are being separated. 


   SERVICES' PLANS FOR REDUCING
   ATTRITION
---------------------------------------------------------- Chapter 2:1

Reducing attrition to zero is neither practical nor possible. 
Attrition will always occur because recruits will have medical
conditions that cannot be discovered in the MEPS examinations, they
will be injured during training, or they will not adapt to military
life.  However, several military officials we spoke with believe that
attrition can be reduced because a portion of it is avoidable.  For
example, some recruits are now being enlisted with medical problems
or with drug use habits that could have been detected earlier in the
enlistment process or that were detected and were waived. 

Though no one in the services can define exactly what portion of
attrition can be avoided, three of the services have set targets for
reducing it.  The Navy, for example, has planned a "War on Attrition"
to reduce attrition at all stages, from recruitment to retention in
the fleet.  The Navy hopes to reduce its attrition at all stages by 5
to 10 percent.  Specifically, the Navy would like to reduce attrition
from basic training by 2,000 persons per year. 

The Army has recently contracted for a study of what an "acceptable"
level of attrition should be.  In the absence of such a defined
level, the Army has suggested a 4-percent goal for reducing attrition
up to the 6-month point. 

The Marine Corps has recently proposed several initiatives to reduce
enlisted attrition at various stages of the training pipeline. 
However, it has not defined quantitative goals for reducing
attrition. 

Finally, the Air Force has taken a new look at enlisted attrition. 
In December 1995, the Air Force began to look at issues that pertain
to military attrition.  According to Air Force officials, the Air
Force's fiscal year 1997 budget proposal contains goals for reducing
attrition.  The Air Force has accordingly reduced its budget on the
assumption that it will be able to reduce its current basic training
attrition rate from 9.5 percent to 7 percent and its first-term
attrition after basic training by 5 percent. 


   THE SERVICES MAKE A SUBSTANTIAL
   INVESTMENT IN RECRUITS WHO
   SEPARATE IN THE FIRST 6 MONTHS
---------------------------------------------------------- Chapter 2:2

The military services' investment in their enlisted personnel is made
up of both fixed and variable costs.  The fixed costs can be thought
of as overhead or infrastructure costs that are not easily or quickly
changed and cannot be directly associated with a single enlistee. 
Examples of this type of cost are the total number of recruiters or
drill instructors or the money spent by a service on a television
advertisement campaign for recruiting.  The variable costs are
directly connected to each recruit, such as costs for MEPS
examinations, transportation from MEPS to basic training, issuance of
clothing, and pay and allowances for each enlistee. 

On the basis of DOD cost data, we estimate that in fiscal year 1996,
DOD and the services spent about $390 million in fixed and variable
costs to recruit and train individuals who never made it to their
first duty stations.  It costs between $9,400 and $13,500 to recruit
and train an active-duty enlistee through basic training and an
additional $6,100 to $16,300 to train the enlistee in an initial
skill. 

To calculate the services' investment in enlistees who separated in
fiscal year 1994, we multiplied the numbers of those separated at the
2-month and 6-month points by the average investment per enlistee. 
We chose the 2-month point because at this time, most recruits have
completed basic training.  We chose the 6-month point because by that
time, most recruits have completed follow-on training.  We used the
most current attrition cost figures available--for fiscal year 1993. 
We converted fiscal year 1993 dollars to fiscal year 1996 dollars. 

Of the services' $390 million investment in enlistees who never made
it to their first duty stations, about 60 percent of this investment,
or $231.8 million, was made in enlistees who were separated in their
first 2 months of service (see table 2.1). 



                               Table 2.1
                
                  Services' Investment in Recruits Who
                 Enlisted in Fiscal Year 1994 and Were
                    Separated in the First 2 Months

                       (Fiscal year 1996 dollars)

                                          Number  Investme
                                              of     nt in       Total
                        Number  Attritio  attrit      each  investment
                            of    n rate      ed  separate      in all
                      accessio  (percent  enlist         d   separated
Service                     ns         )     ees  enlistee   enlistees
--------------------  --------  --------  ------  --------  ----------
Army                    61,408      9.85   6,051   $13,522  $81,821,62
                                                                     2
Navy                    53,501     12.56   6,721    12,077  81,169,517
Marine Corps            31,759      9.81   3,114    14,322  44,629,848
Air Force               29,760      8.69   2,585     9,360  24,195,600
======================================================================
Total                  176,428     10.47  18,471            $231,816,5
                                                                    87
----------------------------------------------------------------------
Source:  GAO analysis of DOD and DMDC data. 

About 40 percent of the services' investment in enlistees who were
separated in the first 3 to 6 months, or $158.3 million, was made in
enlistees who were discharged between the 3rd and 6th months of
service (see table 2.2). 



                               Table 2.2
                
                  Services' Investment in Recruits Who
                 Enlisted in Fiscal Year 1994 and Were
                  Separated in the First 3 to 6 Months

                       (Fiscal year 1996 dollars)

                                          Number  Investme
                                              of     nt in       Total
                        Number  Attritio  attrit      each  investment
                            of    n rate      ed  separate      in all
                      accessio  (percent  enlist         d   separated
Service                     ns         )     ees  enlistee   enlistees
--------------------  --------  --------  ------  --------  ----------
Army                    61,408      5.69   3,493   $20,733  $72,420,36
                                                                     9
Navy                    53,501      3.22   1,723    26,552  45,749,096
Marine Corps            31,759      2.78     884    20,426  18,056,584
Air Force               29,760      2.89     859    25,672  22,052,248
======================================================================
Total                  176,428      3.94   6,959            $158,278,2
                                                                    97
----------------------------------------------------------------------
Source:  GAO analysis of DOD and DMDC data. 


   SHORT-TERM SAVINGS IN VARIABLE
   COSTS BY REDUCING ATTRITION
---------------------------------------------------------- Chapter 2:3

Significant near-term savings in variable costs could result from
screening out the applicants who are now enlisting and are almost
immediately being separated.  For example, if recruiters send
individuals with medical disqualifications to the MEPS, the service
still pays for a MEPS examination, which costs around $70.  If the
individuals with these disqualifying medical conditions make it
through their medical examinations, the services must pay for their
transportation to basic training and then pay, clothe, house, and
feed these recruits while they await separation.  After separation,
the services must pay to transport the enlistees home.  As another
example, when the services do not test for drugs until the recruits
arrive at basic training, those services incur all the marginal
costs, which could have been avoided had the services tested the
recruits for drugs at the MEPS.\1 If the services have to add
screening mechanisms in order to disqualify recruits earlier, the
cost of these additional mechanisms would have to be subtracted from
any calculations of marginal savings.  Such added screening
mechanisms could include requiring more preenlistment documentation
or medical tests. 

Marginal cost savings resulting from improved and earlier screening
of recruits could be realized immediately.  The marginal cost of
sending a recruit to basic training and then separating him or her
can be substantial.  For example, the Navy calculates that its
marginal cost for each recruit who is separated from basic training
is $4,700 for each male and $4,900 for each female.  These figures
are based on the Navy's estimate that it costs $83 to transport a
recruit to basic training; $3,650 to pay, feed, and house the recruit
while at basic training;\2 $91 to provide the recruit's medical
examination at basic training; $817 to provide a male recruit with
clothing ($995 for a female recruit); and an additional $83 to
transport the recruit home after separation.  If the Navy were to
screen out recruits for medical or drug disqualifications after the
MEPS examination but before sending them to basic training, its
immediate cost savings would be at least $4,700 per recruit.\3

Assuming that the Navy's marginal costs are comparable to those of
the other services, we estimate that the marginal cost savings
realized through a 4-percent reduction of attrition would be $4.8
million.  With a 10-percent reduction in 6-month attrition, the
services could realize $12 million in savings.  (See table 2.3.)



                               Table 2.3
                
                 Marginal Cost Savings to Be Gained by
                 Reducing Attrition by 4 and 10 Percent

                 (Fiscal year 1996 dollars in millions)

                        Savings resulting from  Savings resulting from
                         a 4-percent reduction  a 10-percent reduction
Service                           in attrition            in attrition
----------------------  ----------------------  ----------------------
Army                                      $1.8                    $4.5
Navy                                       1.6                     4.0
Marine Corps                               0.8                     1.9
Air Force                                  0.6                     1.6
======================================================================
Total                                     $4.8                   $12.0
----------------------------------------------------------------------
Source:  GAO's analysis of DOD and DMDC data. 


--------------------
\1 The marginal cost is the variable cost of recruiting and training
each recruit. 

\2 This calculation is based on the Navy's estimate that the average
recruit remains at basic training 25 days before being separated and
costs the Navy $146 per day. 

\3 We requested similar cost data from the other three services. 
They were unable, however, to provide us with marginal costs
comparable to those of the Navy because (1) the services'
methodologies in calculating costs differed, (2) the services
captured different data elements, and (3) the services did not
capture certain data elements that are necessary to calculate how
much it costs to send recruits to basic training and then separate
them.  For example, the Marine Corps did not track the average time
in service of an enlistee who is separated during basic training. 
Data provided to us by the Army did not distinguish between fixed and
variable costs, and the Army's average cost was calculated using the
cost of all enlistees who separate during their first terms. 
Finally, the Air Force provided us with the variable cost per
graduate from basic training, but not the cost of each separated
enlistee. 


   POSSIBILITY OF LONG-TERM
   SAVINGS THROUGH INFRASTRUCTURE
   CUTS
---------------------------------------------------------- Chapter 2:4

Over time, if the number of unqualified recruits were significantly
reduced through better screening, it would be possible to reduce the
services' infrastructure associated with recruiting and training,
resulting in savings due to lower fixed costs.  An important caveat
is that these cost reductions probably would not be proportional to
the decrease in attrition.  For example, if attrition were reduced by
10 percent, it is likely that infrastructure costs would fall by
something less than 10 percent.  One important reason that
infrastructure costs are not likely to decrease in the same
proportion as attrition falls is that the services may need to ensure
that their recruiting and training organizations maintain excess
capacity in the event of future increases of accessions.  The
services now determine staffing and funding for recruiting commands
based on the services' accession missions, which have the potential
for being lower if attrition were to decrease. 

Despite these caveats, we believe that it provides perspective to
demonstrate the magnitude of the possible savings to be gained
through reducing attrition and the associated recruiting and training
infrastructure.  To provide this perspective, we have chosen the
services' highest and lowest attrition goals:  a 4-percent reduction
and a 10-percent reduction of 6-month attrition (see table 2.4). 



                               Table 2.4
                
                   Long-Term Savings by Achieving 4-
                Percent and 10-Percent Reductions in 6-
                            Month Attrition

                 (Fiscal year 1996 dollars in millions)

                        Savings resulting from  Savings resulting from
                         a 4-percent reduction  a 10-percent reduction
Service                           in attrition            in attrition
----------------------  ----------------------  ----------------------
Army                                      $6.2                   $15.4
Navy                                       5.1                    12.7
Marine Corps                               2.5                     6.3
Air Force                                  1.8                     4.6
======================================================================
Total                                    $15.6                   $39.0
----------------------------------------------------------------------
Source:  GAO's analysis of DOD and DMDC data. 


   DANGERS OF ESTABLISHING
   ATTRITION TARGETS WITHOUT
   ADEQUATE INFORMATION ON WHY
   RECRUITS ARE SEPARATED
---------------------------------------------------------- Chapter 2:5

We agree that reducing attrition is possible and that the services'
current targets for reducing attrition may represent modest and
achievable goals.  However, the services do not know whether more
reductions are possible.  DOD and the services do not currently have
adequate information to determine how much attrition is avoidable and
therefore should be cut.  Establishing arbitrarily defined targets
for reducing attrition, without knowing precisely what these targets
should be, could result in the services' retaining less qualified
recruits. 

According to officials throughout the services, reducing attrition
would be no problem.  They feared, however, that cutting attrition
could result in a corresponding reduction in the quality of their
enlistees.  That is, service officials feared that limiting attrition
could force them to retain less qualified recruits.  In 1980, we also
anticipated this possible negative effect of attrition ceilings.\4 At
that time, we expressed concern that the Office of the Secretary of
Defense and the services "might, because of congressional concern
over attrition levels, attempt to control rather than manage
attrition." We stated that "While control, through such means as
attrition ceilings, is a quick and easy way to reduce attrition, it
could ultimately prove counterproductive by retaining in the service
persons who do not belong there, which would result in equally
serious problems." We also pointed out that before DOD can
effectively manage attrition, it must have adequate data on the
reasons that enlistees separate early. 


--------------------
\4 Attrition in the Military--An Issue Needing Management Attention
(GAO/FPCD-80-10, Feb.  20, 1980). 


   DOD DOES NOT HAVE DATA
   AVAILABLE TO ESTABLISH
   APPROPRIATE TARGETS FOR
   REDUCING ATTRITION
---------------------------------------------------------- Chapter 2:6

In 1980, as during this review, we found that DOD did not have data
on attrition that allowed it to assess service-wide attrition trends
and the factors behind their changes.  DOD's data is inconsistent and
incomplete for two reasons.  First, DOD's primary source of
service-wide attrition data--which is managed by DMDC--contains only
the officially assigned separation codes assigned to enlistees, when
many other reasons may drive enlistees' discharges.  Second, the
services interpret DOD's definitions of the separation codes
differently and therefore place enlistees with identical situations
in different discharge categories.  Because of these two drawbacks,
DMDC's attrition data can be used to deduce only a rough estimate of
why enlisted personnel leave the services. 


      DOD'S PRIMARY DATABASE
      CAPTURES ONLY THE OFFICIAL
      REASON FOR AN ENLISTEE'S
      SEPARATION
-------------------------------------------------------- Chapter 2:6.1

DMDC data captures only one of many possible reasons that enlistees
leave the service.  The reasons for separation that are collected in
DMDC's database are based on separation codes taken from a
servicemember's official discharge form, the DD Form 214.\5 The
separation program designator is a three-character code that captures
the service's official reason for separation.  DMDC converts these
designators into interservice separation codes, which it developed in
an attempt to enable cross-service comparisons of separation reasons. 

Our analysis of these separation codes and our interviews with
service officials and over 100 separating recruits revealed that
enlistees generally have many reasons for leaving, only one of which
is recorded in DMDC's database.  A 1991 Rand study of enlisted
personnel files also found that over 80 percent of the recruits whose
files they examined had multiple reasons for their early release.\6
Rand found that, typically, the separation code chosen by a service
to go on the servicemember's DD-214 "is the separation code that the
service believes would provide the most direct path to a successful
discharge or that would offer the strongest legal case.  It does not
indicate the actual reason why the recruit separated early."


--------------------
\5 The DD Form 214 is a servicemember's "Certificate of Release or
Discharge From Active Duty."

\6 Why Recruits Separate Early (R-3980-FMP, 1991). 


      SERVICES INTERPRET DOD'S
      DEFINITIONS OF SEPARATION
      CODES DIFFERENTLY
-------------------------------------------------------- Chapter 2:6.2

In our 1980 report, we recommended that DOD improve existing
management information systems to include attrition data-reporting
systems that were more uniform.  At that time, as at present, each
service had designed its own system of classifying attrition by
reasons, and these systems varied by installation, according to
commanders' interpretations of criteria.  Consequently, DOD and the
services were not able to compare trends by cause among the services. 

In an attempt to standardize the services' use of these codes, DOD
issued a list of the codes with their definitions.  However, it has
not issued implementing guidance for interpreting these definitions,
and the services' own implementing guidance differs on several
points.  While we believe that the individual services and local
commanders should have flexibility in managing their personnel
programs, we also believe that DOD and service headquarters have the
responsibility for ensuring that their separations of enlisted
personnel are consistent with overall DOD policy, are uniformly
applied, and are effective. 

Assigning a particular separation code rather than another has many
implications.  As we were told by service officials, it is in the
best interest of basic training personnel to assign separation codes
that reflect least poorly on the basic training site.  For example,
some officials told us that if a recruit has minor medical problems
but even more severe motivational problems, basic training personnel
might choose to separate this recruit for medical reasons, as opposed
to performance problems.  Separating a recruit for medical rather
than performance problems is face-saving for both the recruit and
basic training personnel.  The recruit does not have to admit that he
or she could not meet the minimum performance standards, and the
command does not have to admit that it could not motivate the
recruit. 

Conversely, Army officials we spoke with said that, until recently,
one Army basic training site had been discharging injured recruits
for failing to meet minimum performance standards.  They also said
that, since personnel at the Army's basic training site had been told
by upper-level officials that its separations for preexisting medical
conditions were too high, they expected that the training command
would begin to separate recruits with medical problems under other
codes. 

During our review, we found the following examples, among others, of
the differences in the way the services assign separation codes: 

  -- An enlisted person who exhibits a situational adjustment problem
     in adapting to military life is separated from the Air Force for
     a personality disorder, from the Navy for an erroneous
     enlistment, and from the Marine Corps for failure to meet
     minimum performance standards.  Air Force personnel told us that
     most persons separated for adjustment disorders do not have true
     personality disorders, and the discharge documentation could
     cause persons separated with this description to carry a stigma
     into later life. 

  -- An enlisted person who is discharged for a disqualifying medical
     condition that he or she did not know about in advance may be
     discharged from the Marine Corps under the separation code for
     the convenience of the government or for erroneous enlistment. 
     In the Army, this same person is separated using the separation
     code indicating that he or she did not meet medical/physical
     standards. 

  -- If an enlistee intentionally withholds medical information that
     would disqualify him or her and is then separated for this same
     medical condition, the enlistee is discharged from the Air Force
     and the Marine Corps for a fraudulent enlistment.  The Army
     categorizes this separation as a failure to meet
     medical/physical standards unless it can prove that the enlistee
     withheld medical information with the intent of gaining
     benefits.  The Air Force and the Marine Corps do not require
     this proof of intent.  The Navy categorizes this separation as
     an erroneous enlistment, which indicates no fault on the part of
     the enlistee. 


      DMDC DATA PROVIDES ONLY A
      ROUGH ESTIMATE OF MAJOR
      REASONS FOR ATTRITION
-------------------------------------------------------- Chapter 2:6.3

DMDC data for fiscal year 1994 shows that DOD's attrition rate was
14.4 percent at the 6-month point in enlistees' first terms.  This
means that in fiscal year 1994, 25,430 enlisted personnel were
separated from the services within the first 6 months of their
enlistment terms.  Of this number, 21,229, or about 83 percent, were
assigned separation codes indicating that they (1) were medically
unqualified for military service, (2) demonstrated character or
behavior disorders, (3) fraudulently or erroneously entered the
military, or (4) failed to meet minimum performance criteria. 

While these separation codes represent general areas in which the
services suffer the most early attrition, the codes cannot provide a
basis for determining how much of the attrition in these areas can be
reduced.  For example, in the area of medical disqualifications,
DMDC's data does not quantify what percentage of enlisted personnel
discharged with this separation code had verifiable medical
conditions that could have been screened out in advance.  In the area
of character and behavior disorders, the data does not distinguish
between those separated for severe personality disorders and those
who experienced mild situational disorders and might have been having
motivational problems.  In the area of fraudulent and erroneous
enlistment, the data does not allow the services to determine what
percentage of those separated in this category were discharged for
concealing criminal background histories, as opposed to medical or
psychological conditions.  Finally, in the area of performance
separations, the data does not allow DOD to determine what percentage
of these enlistees were separated for minor injuries or what
percentage might have been further counseled or rehabilitated. 


      DOD HAS PROPOSED A PROJECT
      TO DEVELOP A DATABASE ON
      MEDICAL DISCHARGES
-------------------------------------------------------- Chapter 2:6.4

According to an official in DOD's Office of the Assistant Secretary
of Defense for Health Affairs, DOD approved a project in July 1996 to
compile a comprehensive database of medical conditions for all
accessions.  This database project was funded in September 1996.  As
part of the project, doctors throughout the services will be required
to use an internationally recognized medical code book to diagnose
medical conditions for all their patients.  This information will be
fed into a database maintained by DMDC and will then be analyzed by
the Walter Reed Institute of Research.  This database will provide
DOD with the ability to reevaluate its physical enlistment standards,
to analyze the medical reasons that recruits are separated, and to
make fact-based policy changes to reduce medical attrition. 


   RECOMMENDATION
---------------------------------------------------------- Chapter 2:7

To provide a reliable database for DOD to manage attrition and for
the services to set appropriate targets for reducing attrition, we
recommend that the Secretary of Defense issue implementing guidance
for DOD's separation codes. 


   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 2:8

In commenting on a draft of this report, DOD partially concurred with
our recommendation that the Secretary of Defense issue implementing
guidance for DOD's separation codes (see app.  I).  DOD agreed that
"the administrative techniques at the user level for coding these
losses require attention in order to validate attrition rates and the
recommendation to enforce standardization is valid." DOD also
stressed that its separation codes were the subject of a
standardization initiative in 1993, which provided a common list to
all services for use starting October 1, 1993.  As we state in our
report, the lack of standardization in the services' use of
separation codes is a long-standing issue, dating back to at least
1980.  While DOD's issuance of a list of standardized separation
codes was a step in the right direction, we found that the services
have not been applying these codes consistently and that this lack of
consistency makes it impossible for DOD to analyze reasons for
attrition on a service-wide basis. 

DOD questions our statement that basic training personnel sometimes
choose the separation code that allows them to "save face," stating
that the original intent of the separation codes listed on the
DD-214s was not to capture reasons for attrition.  Though this may be
the case, these codes are at present the only service-wide
information on reasons for separations.  If DOD decides that it would
be more feasible to collect service-wide reasons for separation using
a different mechanism, we would not object.  Our concern is that such
data be collected and analyzed, not that the data be based on
separation codes listed on the DD-214s. 


THOUSANDS OF RECRUITS ARE
SEPARATED EARLY BECAUSE THEY ARE
UNQUALIFIED OR UNMOTIVATED
============================================================ Chapter 3

During this review, we focused on separations in three categories: 
separations for medical conditions that were not disabilities,
separations for drug use, and separations for failure to meet
performance standards.  Separations for medical conditions and
failure to meet performance standards represent at least 55 percent
of all 6-month attrition for enlistees who entered the services in
fiscal year 1994.  This percentage is understated for two reasons. 
First, some persons who have medical problems are being separated as
fraudulent enlistments because they concealed their medical
histories.  Second, some persons who have performance problems are
being separated for character or behavior disorders.  We were not
able to calculate the number of persons discharged for drug use in
all the services because separations for these persons are
categorized in many different ways.  For example, a person who uses
drugs could be separated for erroneous or fraudulent enlistment,
personality disorder (if found to be drug-dependent), misconduct, or
drug rehabilitation failure. 

A calculation of the numbers of persons in each of these three
categories can only be approximate because the services assign
separation codes differently and because persons placed in one of
these categories might have actually been separated for many
different reasons.  Even so, the data indicates in a general way why
attrition during the first 6 months of an enlistee's term occurs. 
This data--along with our interviews with recruiting, examining, and
training personnel and with separating recruits--indicates that the
services' screening processes are not identifying significant numbers
of persons who have disqualifying medical conditions or who use
drugs. 

The data also indicates that DOD separates thousands of personnel who
do not meet minimum performance standards.  Recruits have problems
meeting performance standards because they are not physically
prepared for basic training and because they lack motivation.  GAO's
interviews with separating recruits and recent research suggest that
negative leadership techniques at basic training may contribute to
some recruits' lack of motivation to meet performance standards. 


   SCREENING PROCESSES DO NOT
   IDENTIFY THOUSANDS OF RECRUITS
   WHO ARE UNQUALIFIED FOR SERVICE
---------------------------------------------------------- Chapter 3:1

DOD's current processes for screening enlisted recruits are
inadequate in the following ways: 

  -- Recruiters do not have sufficient incentives to screen out
     persons who may not be fully qualified to complete basic
     training. 

  -- Recruiting and MEPS screening mechanisms and service waiver
     policies result in the enlistment of thousands of persons who
     have preexisting medical conditions and who are later separated. 
     Also, the responsibility for reviewing cases involving
     preexisting medical conditions resides with MEPCOM, which poses
     a conflict of interest because this is the command most directly
     responsible for determining the physical qualifications of
     military applicants. 

  -- The Navy and the Marine Corps do not test recruits for drugs
     until they arrive at basic training. 


      RECRUITERS' INCENTIVES ARE
      NOT ADEQUATELY TIED TO
      ENLISTEES' SUCCESSFUL
      COMPLETION OF BASIC TRAINING
-------------------------------------------------------- Chapter 3:1.1

In a sense, recruiters have a conflict of interest.  Though they are
obliged to their services to recruit only qualified personnel, their
performance is judged primarily on how many recruits they enlist per
month.  Recruiters' monthly recruiting goals are established on the
basis of the services' accession needs, which are in turn driven by
end strength numbers and budget allocations.  The recruiters' goals
are also connected to the numbers of basic and follow-on training
slots.  That is, recruiters must keep a steady and constant flow of
enlisted personnel into the services.  At the same time, recruiters
are the first step in the process of determining whether applicants
are qualified mentally and physically to serve in the military. 
Despite this secondary but very important function, recruiters
receive no credit for screening out unqualified applicants before
they are enlisted. 

Recruiters' quotas are tied to whether their recruits enter active
duty after being in the Delayed Entry Program.  After enlisting,
recruits may remain in the Delayed Entry Program from 2 days to 1
year.  If an enlistee drops out of the Delayed Entry Program, the
recruiter gets no credit for that enlistment.  After the recruit
begins active duty and is transported to basic training, only two
services--the Marine Corps and the Navy--continue to hold their
recruiters responsible for whether their recruits successfully
complete basic training. 

The Air Force and the Army do not tie their recruiters' incentive
systems to successful completion of basic training because they
believe that recruiters should not be penalized for their recruits'
failure to complete basic training.  According to Air Force
officials, the Air Force provides feedback to its recruiters on each
loss from training and produces data quarterly to identify recruiters
whose recruits have high or low training attrition rates.  This
allows Air Force supervisors to take corrective actions with
recruiters whose recruits are separated at a high rate.  The Army
believes that once applicants are transported to basic training,
recruiters have done their jobs and are no longer able to influence
their enlistees' performance. 

Beginning in June 1996, the Navy implemented a revised recruiter
incentive system that will subtract points from recruiters whose
enlistees do not complete basic training.  Because the new system has
just been implemented, the Navy has not yet determined whether the
system will help to reduce attrition. 

For many years, the Marine Corps has given its recruiting units the
flexibility to design incentive systems that factor in successful
basic training graduation rates and to deduct recruiter points when a
recruit fails basic training.  According to Marine Corps recruiting
officials, however, their recruiting incentive systems are not
uniform because Marine Corps headquarters authorizes each recruiting
district and station to develop its own awards systems.  That is,
some recruiting stations may choose to award additional points to
recruiters for each successful basic training graduate or to take
points away for each recruit who drops out, and others may not.  In
any case, some Marine Corps basic training officials suggested that
Marine Corps recruiters are still driven primarily by their monthly
goals.  These officials said that the threat that recruiters might
lose points 3 months later (when the recruit fails basic training)
does not provide sufficient motivation for Marine Corps recruiters to
do more thorough screening. 

Basic training officials from all services told us that they believed
that recruiters do not have adequate incentives to ensure that their
recruits are qualified medically, morally, and psychologically.  That
is, these officials believe that recruiters are driven by their
monthly goals to recruit persons who may not be fully qualified and
that recruiters do not have an incentive to thoroughly probe
applicants to learn of possibly disqualifying medical, psychological,
or criminal problems. 

Recruiters are not solely responsible for the services' failure to
screen out unqualified recruits.  We agree with service officials
that it is not the recruiters' job to determine whether recruits are
medically qualified.  Rather, it is the MEPS physicians' job to
perform this function.  Even so, we believe that the services do not
provide recruiters with adequate incentives to ask applicants probing
questions that might reveal disqualifying information.  Asking
probing questions leads to two complications for recruiters.  First,
if recruiters uncover potentially disqualifying information about
their applicants, they create more paperwork for themselves in that
they must request waivers.  Second, recruiters might have to reject
applicants who are not qualified and miss their monthly quotas. 

In some cases, recruiters are given bonus points for recruiting top
quality recruits, "quality" recruits being defined as those who have
high school diplomas and score in the upper percentiles of the ASVAB. 
However, recruiters are not sufficiently rewarded for other
indicators of excellence in enlisting personnel for the services,
measures such as thoroughness of screening, suitability of matches
between an applicant and the job for which he or she is most
qualified, and length of time the applicant stays in the military. 
These indicators suggest that recruiters' incentive systems could be
improved by comparing them to incentive systems for job counselors. 
Recruiters are trained to be sales people, rather than job counselors
or placement officers. 

Two suggestions were made to us during our fieldwork for improving
recruiters' incentive systems to relieve the pressure to recruit
unqualified personnel.  First, it was suggested that recruiters
receive partial credit for thoroughly probing applicants and
ultimately finding that they are unqualified.  This partial credit
would provide an incentive for recruiters to fully screen applicants
without losing all credit for investing time in working with these
persons. 

A second idea for improving the recruiter incentive systems was to
give recruiters a "floating goal." Under this type of system, a
recruiter's goal would not be set for one month only but could, for
example, be a 3-month goal that continued to move forward.  With such
a floating goal system, a recruiter would be provided with an
incentive to overproduce one month to relieve the pressure of the
next month's goal.  Conversely, if the recruiter were unable to meet
one month's goal, he or she would be able to make up that lack in the
next month.  According to officials from the Office of the Secretary
of Defense, this floating goal system has been tried and has met with
failure.  Past experience showed that recruiters tended to wait until
the last possible moment to meet their recruiting goals.  While we
did not evaluate why this floating goal system did not work in the
past, we continue to believe that innovative ideas like these merit
further consideration and study. 


      THOUSANDS ARE SEPARATED FOR
      PREEXISTING MEDICAL
      DISQUALIFICATIONS
-------------------------------------------------------- Chapter 3:1.2

The services are enlisting persons with disqualifying medical
conditions for two primary reasons:  (1) applicants conceal their
medical histories and (2) the services waive medical conditions that,
according to DOD directives, are disqualifying.  The Military
Entrance Processing Command reviews the personnel files for all
enlistees who are separated for preexisting medical conditions to
determine when medical screening has not been effective.  However,
the fact that the responsibility for reviewing these cases lies with
the same command that does the medical screening raises the
possibility of a conflict of interest. 

DMDC data indicates that approximately 6,800 of the nearly
176,400 enlistees who entered the services in fiscal year 1994, or
3.9 percent, were found to be not medically qualified for service
within the first 6 months of their terms.  Of the approximately 6,800
personnel who were separated for preexisting medical conditions,
around 3,600 were in the Army, 1,700 were in the Navy, 1,400 were in
the Air Force, and 100 were in the Marine Corps.  The Air Force's and
the Marine Corps' numbers are understated because they separate
persons who conceal medically disqualifying information for fraud. 

The problem with enlisting medically unqualified recruits into the
services is a long-standing one.  In 1965, in 1968, and again in
1970, we reported the numbers of personnel who enlisted in the
services with preexisting medical conditions.\1 We estimated that
$17.9 million was expended during fiscal year 1969 by the military
services, primarily for pay and allowances, uniforms, and travel, for
personnel discharged because of preservice physical defects after
serving in the military for a year or less. 

In our 1970 report, we stated that discharges for medical conditions
that existed prior to service accounted for 2.3 percent of all
accessions between fiscal years 1966 through 1969.  During our
current review, MEPCOM reported that discharges in this category
represented 2.9 percent of all accessions in fiscal year 1994 and 3.3
percent of all accessions in fiscal year 1995.  According to an
official in the Office of the Secretary of Defense, enlistees today
are more likely to conceal disqualifying medical conditions than they
were during the draft.  During the draft, many inductees wished to
avoid military service, while now enlistees are voluntarily enlisting
and have more of an incentive to conceal medical conditions. 


--------------------
\1 Unnecessary Costs Incurred Because of Acceptance of Physically
Unqualified Enlisted Members in the Armed Services (B-14686, Apr.  2,
1965); Report on Matters Relating to Enlisted Personnel Discharged
Because of Defects That Existed Prior to Entrance Into Military
Service (B-146986, Mar.  21, 1968); and Discharge of Military
Enlisted Personnel Because of Preservice Physical Defects (B-146986,
July 27, 1970). 


         APPLICANTS ARE NOT
         SYSTEMATICALLY REQUIRED
         TO PROVIDE MEDICAL
         RECORDS PRIOR TO
         ENLISTMENT
------------------------------------------------------ Chapter 3:1.2.1

MEPCOM's data for fiscal year 1994 indicates that over half of all
separations for preexisting medical conditions involved the
applicants' concealment of their medical conditions.  Concealment of
past medical history is made easier by the fact that applicants are
required to present medical histories only if they report past
medical problems.  Applicants who wish to join the service have an
incentive to conceal such information. 

Of all the services, the Navy had the most complete data on whether
enlistees who were separated for medical conditions had known about
and reported these conditions prior to enlistment.  Personnel at the
Navy's basic training clinic administer a questionnaire to all
recruits who are separated for preexisting medical conditions asking
them whether they had previously reported their conditions. 
According to Navy clinic personnel, 1,684 Navy recruits were
separated for medical conditions in calendar year 1995.  Of these, 43
percent, or 724 recruits, reported on their questionnaires that they
had seen civilian physicians at some time before they enlisted in the
Navy about the condition for which they were separated.  Fifty-five
percent of all those discharged for preexisting medical conditions
(928 recruits) reported that they had told their recruiters of their
conditions before enlisting, and 41 percent
(683 recruits) reported that their recruiters had told them not to
mention the medical conditions.  Navy clinic personnel told us that
"identifying those patients with conditions not compatible with Navy
service prior to their enlistment saves training dollars and keeps
these patients away from a potentially dangerous environment."
However, while Navy doctors had compiled these statistics, they knew
of no resulting changes to the enlistment screening process. 

According to Navy recruiting officials, data compiled by Navy doctors
from this questionnaire began to be reported to the Navy's Recruiting
Command in January 1996.  These officials said that to date, 60
percent of all the fiscal year 1997 allegations against recruiters
had been immediately retracted by recruits.  The remaining 40 percent
of these allegations were referred to the Navy Recruiting Command's
Inspector General for further investigation. 

Personnel from basic training camps for all four services told us
that they share information with their recruiting commands on what
enlistees tell them about recruiters' failure to encourage the
enlistees to completely divulge medical conditions.  However, they
also stressed that enlistees often blame their recruiters when they
themselves are guilty of withholding medical information.  According
to these officials, only a very small percentage of enlistees'
allegations about recruiters are substantiated.  These officials also
told us that when it comes down to the recruit's word against the
recruiter, the services side with the recruiter. 

Personnel throughout the services stressed the difficulty of
encouraging applicants to be honest in divulging their past medical
conditions and the difficulty of obtaining past medical records.  If
an applicant has lived in several different places, it would be
extremely difficult for the services to research medical records from
dozens of hospitals. 

The services now ask applicants to provide medical records only when
they divulge past medical problems.  Neither do the services ask all
applicants to provide the names of their medical insurers and medical
providers.  Asking all applicants to provide such information might
reduce the number who do not fully disclose past medical histories. 
We realize that some applicants may not be insured; some might have
had several different insurers and visited many hospitals and doctors
over a period of years; and others may continue to withhold
information about their medical histories.  However, we believe that
applicants would be less likely to conceal past medical problems if
they were required to (1) provide the names of their medical insurers
and past medical providers and (2) sign a release form allowing the
services to request their medical records. 

The services would not necessarily have to obtain medical records for
all applicants.  Rather, the services could determine how frequently
they needed to research applicants' medical histories in order to
ensure that applicants believed that the information they provided
the services could and might be verified.  To ensure that the
paperwork burden of recruiters was not increased and that recruiters
themselves did not know which applicants would receive this greater
scrutiny, the services could place the responsibility for obtaining
medical files on persons who are not production recruiters. 

Officials from the Office of the Army's Deputy Chief of Staff for
Personnel suggested that one reason that applicants may not be
reporting disqualifying medical information is that the questions on
the "Applicant Medical Prescreening Form" (DD Form 2246) are not
specific enough.  Recruiters use this form to ask applicants about
their past medical histories and drug use.  Army officials believe
that recruiters are following "the letter of the law" when gathering
medical data.  That is, recruiters are asking the questions contained
on the form verbatim and are not asking follow-up questions at least
in part because they are not medically qualified to know what the
relevant medical questions are.  Army officials suggested that the
prescreening form might be revised to make the questions more
specific and to tie the questions more closely to medical conditions
that most often result in recruits' separations. 

During our review, we also found that some of the questions on the
recruiter's prescreening form were vague and ambiguous.  For example,
one question is "Have you ever had or have you now back trouble?" The
form does not define or give examples of what an applicant should
consider to be "back trouble." Another question is "Have you ever had
or have you now addiction to drugs or alcohol?" There is no question
on the form asking whether the applicant has tried alcohol or drugs
or how often the applicant has used alcohol or a particular drug. 

After applicants have completed the recruiter's prescreening form,
they fill out a second medical history form at the MEPS, the "Report
of Medical History" (SF 93).  During our visit to the Navy's basic
training camp, doctors there said that they believed that the MEPS
medical history form was obsolete, ambiguous, and easy for applicants
for falsify.  For example, one of the questions on the form is "Have
you ever had or have you now lameness?" Navy doctors pointed out that
the term "lameness" is no longer used by medical personnel.  The form
also asks, "Have you ever had or have you now [a] 'trick' or locked
knee?" The form contains no definitions of "trick" or "locked." Also,
Navy basic training doctors said that they could tell that applicants
do not thoroughly read the form as they record their answers.  These
doctors believe that applicants often automatically check "no" to all
questions, not realizing that one question requires a "yes" answer if
an applicant is to be determined qualified.  The question that
requires a "yes" answer asks whether applicants have vision in both
eyes.  Navy doctors often find that applicants have checked "no" to
this question and then have had to go back to correct and initial
their answer. 

In an effort to get more complete and accurate information from
recruits on their medical histories, the Navy's basic training
doctors ask recruits to fill out an automated form upon arrival at
basic training.  This questionnaire, which the Navy doctors wrote
themselves, contains more specific questions and requires recruits to
fully read the questions because some questions require "yes"
answers, and others require "no" answers.  Also, for some questions,
the doctors deliberately use different terminology to ask the same
medical question.  Navy basic training doctors believe that asking
questions in different ways makes it more likely that recruits will
recognize their particular symptoms and report them. 


         DOD DOES NOT HAVE
         EMPIRICAL DATA ON THE
         COST-EFFECTIVENESS OF
         WAIVERS OR MEDICAL
         SCREENING TESTS
------------------------------------------------------ Chapter 3:1.2.2

MEPCOM data for fiscal year 1994 indicates that close to 8 percent of
separations for preexisting medical conditions involved cases in
which the services granted waivers for the very conditions for which
the recruit was later separated.  DOD has set uniform enlistment
standards for all the services in DOD Directive 6130.3, "Physical
Standards for Appointment, Enlistment, and Induction." However, this
same directive also grants the services the authority to waive the
enlistment standards, and, according to MEPCOM officials, the
services frequently do so.  The Army, for example, told us that the
only two medical conditions for which waivers could not be granted
were pregnancy that existed prior to enlistment and human
immunodeficiency virus. 

Personnel throughout the services explained that it is difficult to
determine which medical conditions to waive because some recruits
with disqualifying medical conditions, such as severe flat feet, are
able to successfully complete basic training, while others are not. 
Statistical data is not currently available that would enable the
services to predict which disqualifying medical conditions represent
good attrition risks. 

According to an official in DOD's Office of the Assistant Secretary
of Defense for Health Affairs, DOD's physical enlistment standards
are not empirically linked to performance in the military, but rather
are based on military experience and expert judgment.  This official
cited as an example DOD's change in its asthma policy.  Before
Operation Desert Storm, applicants who had histories of asthma but
had not had an asthma attack since age 12 could qualify for military
service.  Just after the war, however, military leaders recommended
that persons who had any history of asthma should be disqualified. 
This recommendation was based on military leaders' personal
observations of the effects of environmental conditions on
servicemembers during the war. 

DOD's recently approved plan to collect data on the medical histories
of all accessions, including those who separate early, will provide
an opportunity for DOD to base its enlistment standards and waiver
policies on sound research data.  DOD plans to begin revising its
enlistment standards using this new database beginning in fiscal year
1997.  DOD's new database will also allow DOD to determine whether it
would be cost-effective to add screening tests to MEPCOM medical
examinations or to refrain from giving some medical screening tests
at the basic training sites.  At present, some basic training
screening tests are more thorough than those given at the MEPS, and
at other times, recruits are given the same tests at the MEPS and at
basic training.  For example, while some basic training camps do pap
smears, tests for Hepatitis B, and sickle cell anemia, the MEPS do
not.  According to MEPCOM officials, female applicants are not given
pap smears at the MEPS because, statistically, women of this age
typically do not have health problems that would be surfaced through
pap smears.  If this were the case, however, it is not clear why
female recruits are given pap smears when they arrive at basic
training.  When recruits are found to have medical problems related
to these tests given at basic training but not at the MEPS, they are
sometimes separated from the services. 

On the other hand, the MEPS do some examinations that are repeated at
the basic training camps.  For example, all recruits are tested for
vision and hearing at the MEPS, and at basic training, all or
selected recruits are retested.  During our review of separation
files and our interviews with separating recruits, we found many
cases of recruits who were being separated for vision and hearing
problems that were not discovered until the recruits had been
retested at basic training.  MEPCOM officials told us that the vision
and hearing testing done at the MEPS is comparable in sophistication
to the testing done at the basic training camps.  If this is the
case, it is unclear (1) why recruits need to be retested at basic
training and (2) why recruits frequently fail the vision and hearing
tests given at basic training camps after having passed them at the
MEPS. 

Army officials suggested that an alternative to conducting more
thorough medical screening on all recruits would be to conduct more
thorough medical screening only on applicants whose medical histories
indicated problems.  The MEPS already require that applicants who
report medical problems be seen by medical specialists.  However, if
applicants do not report medical problems and MEPS doctors do not
detect problems, applicants can avoid such medical consultations. 
With improved screening forms, with better incentives for recruiters
to collect medical information, and with DOD's proposed medical
database, DOD would have a sound basis for determining which
applicants should undergo more thorough medical examinations. 


         MEPCOM'S REVIEW OF
         MEDICAL CASES POSES A
         POSSIBLE CONFLICT OF
         INTEREST
------------------------------------------------------ Chapter 3:1.2.3

The military services are required to send to MEPCOM all separation
packages for persons discharged for medical conditions that existed
prior to service.  MEPCOM's doctors then categorize the individual
cases to determine whether the medical conditions should have been
detected at the MEPS.  According to MEPCOM's Command Surgeon, in
fiscal year 1994, only 3.3 percent of the medical separation cases
were due to an error on the part of MEPS personnel. 

Basic training doctors and other personnel told us of cases in which
persons arrived at basic training who were very obviously medically
or psychologically disqualified.  They cited cases of a recruit with
a glass eye, numerous recruits who had no vision out of one eye, a
recruit with a hearing aid, recruits with holes in their eardrums,
and a recruit who had severe and debilitating mental problems.  When
we asked MEPCOM officials about these cases, they said that they had
heard similar anecdotal information about recruits who had not been
screened out at the MEPS.  However, they said that, generally,
persons reporting these cases cannot provide the type of information
that would allow MEPCOM to investigate the cases and trace them back
to the appropriate MEPS. 

We reviewed a selected sample of MEPCOM's files on these cases and
its determination of whether medical conditions should have been
detected by MEPS doctors.  We found no instances in which MEPCOM had
improperly classified cases.  However, we believe that there is a
conflict of interest because the responsibility for reviewing and
categorizing cases resides with MEPCOM, which plays the most
significant role in the medical screening process.  MEPCOM's Command
Surgeon told us that the Air Force also categorizes preexisting
medical separation cases and has in certain instances challenged
MEPCOM's classifications. 


      THOUSANDS SEPARATED FOR DRUG
      USE
-------------------------------------------------------- Chapter 3:1.3

The Air Force and the Army test all of their applicants for drugs at
the MEPS.  As a result, they are able to screen out persons who test
positive for drugs at the time of their preenlistment medical
examinations.  The Navy and the Marine Corps, on the other hand, do
not test applicants for drugs at the MEPS but wait until they arrive
at basic training.  The Navy and the Marine Corps then separate many
recruits for testing positive for drug use upon arrival at basic
training. 

After the Army and the Air Force test their recruits for drugs during
the MEPS medical examination, they require recruits to sign a
statement that they will refrain from drug use while in the Delayed
Entry Program.  The Army's and the Air Force's rationale for not
retesting recruits for drugs at basic training is that any drug use
on the part of recruits will demonstrate itself in recruits'
performance. 

In fiscal year 1994, 1,669 recruits were discharged from the Navy
because of drug use.  During our initial visit to the Great Lakes
Naval Training Center, our review of personnel records and interviews
with separating personnel indicated that some of these recruits had
received waivers for preenlistment drug use, and some had been in the
Delayed Entry Program for insufficient periods of time to eliminate
all drugs from their systems.  We reviewed the personnel records of
19 Navy recruits who were separated for drug use.  Of these 19, the
personnel records included documentation that 17 had admitted to drug
use prior to enlistment; 2 had not.  Of those who tested positive for
drugs upon arrival at basic training, seven had been in the Delayed
Entry Program 1 month or less.  Six of these seven had admitted to
prior drug use.  In other words, some recruits were transported to
Navy basic training when recruiters or MEPS personnel knew that they
had used drugs.  Navy recruiting officials said that if these
recruits had admitted to drug use within the past 30 days, they would
not have been sent to basic training.  Since March 1990, this has
been the Navy's policy.  Up until March 27, 1996, the Navy denied
reenlistment rights to enlisted persons who tested positive for drugs
at basic training.  On that date, the Navy changed its policy to
allow some recruits who tested positive for marijuana to reenlist
after a period of 6 months. 

Navy recruit division commanders told us that recruits being
separated for positive drug tests are generally good performers and
want to stay in the Navy.  At the Navy's basic training camp in Great
Lakes, Illinois, we interviewed seven recruits being separated for
drug use.  Four had tested positive for marijuana, two for cocaine,
and one for an hallucinogenic drug.  Of the four who had tested
positive for marijuana, three had been in the Delayed Entry Program 1
month or less.  All four wanted to stay in the Navy, and two wished
to appeal their cases. 

Like the Navy, the Marine Corps tests for drugs at basic training but
not at the MEPS.  While the Navy grants no waivers for those who test
positive for drugs, the Marine Corps grants waivers to enlistees who
test positive for marijuana at basic training.  In fiscal years 1995
and 1996, around 70 percent of recruits who tested positive for
marijuana at Parris Island Marine Corps Recruit Depot were granted
waivers.  The Marine Corps bases its decisions on granting waivers on
(1) what type of drug was used, (2) whether the use was experimental
or routine, (3) whether the drug was used while the enlistee was in
the Delayed Entry Program, (4) whether the enlistee admitted to the
drug use, (5) whether the enlistee is drug dependent, and (6) whether
the enlistee has been convicted of any drug-related offenses. 


   THOUSANDS OF RECRUITS
   DISCHARGED FOR FAILURE TO MEET
   MINIMUM PERFORMANCE CRITERIA
---------------------------------------------------------- Chapter 3:2

Of all recruits entering the services in fiscal year 1994,
approximately 7,200 were discharged in the first 6 months of service
because they failed to meet minimum performance criteria:  around
4,800 in the Army, 1,300 in the Marine Corps, 600 in the Air Force,
and 500 in the Navy.  The Army's numbers for recruits separated in
this category are significantly higher than those of the other
services for at least three reasons:  (1) the Army enlists more
recruits than any other service; (2) the Army places persons with
mild, situational adjustment problems in this category; and (3) until
recently, the Army placed recruits who had been injured in training
in this category.  According to basic training personnel, recruits
who are discharged for failure to meet minimum performance criteria
include those who fail physical training standards, who cannot meet
weight standards, who fail inspections, or who cannot otherwise adapt
to basic training. 


      RECRUITS FAIL TO MEET
      PERFORMANCE STANDARDS
      BECAUSE THEY ARE NOT
      PHYSICALLY FIT
-------------------------------------------------------- Chapter 3:2.1

Basic training personnel throughout the services said that recruits
who are in good physical shape have a greater chance of meeting
overall military performance standards.  Those struggling to meet
physical requirements are often correspondingly demotivated to meet
other military requirements.  Service officials suggested that two
ways of ensuring that recruits are better prepared to succeed in
basic training are to (1) more fully inform them of the physical
training requirements of basic training while they are still in the
Delayed Entry Program and (2) encourage recruits to become physically
fit while they are waiting to go on active duty. 

All of the services encourage enlistees to become physically fit
while they are in the Delayed Entry Program.  The services stress
that these physical activities must be voluntary.  For many years,
the Marine Corps has encouraged its recruiters to provide
opportunities for enlistees to participate in physical training
activities on a voluntary basis.  Recently, the Navy and the Army
have also restructured their Delayed Entry Programs to encourage
recruits to become more physically fit before they enter basic
training.  As part of the Army's new proposed program, which was near
the implementation stages in December 1996, the Army will allow
recruits in the Delayed Entry Program access to military fitness
centers before they report to basic training and reward them with
retirement points for participating in voluntary physical training
exercises with their recruiters.\2 As another part of the Army's
proposed new Delayed Entry Program, enlistees will be required to
learn basic military customs and rules.  Learning such material
before arriving at basic training will lessen recruits' stress in
attempting to learn large amounts of material in the short time
allotted to basic training.  Air Force officials told us that they do
not require recruiters to conduct any formalized physical training
for recruits while they are in the Delayed Entry Program because they
believe that the Air Force would be legally liable for injuries
sustained by recruits while participating in such training. 

All four services stop short of making physical training activities
mandatory for enlistees because they believe that requiring physical
training activities could make the services liable for any injuries
enlistees sustain while in such training.  Yet, according to Army
legal officials, enlistees are already entitled to the use of
military health care facilities when they are participating in
training activities.  Enlistees in the Delayed Entry Program are
members of the Individual Ready Reserve in an inactive duty status. 
We believe that taking more assertive steps to make recruits
physically fit will result in fewer injuries during basic training
and in reducing the number of enlistees who are separated during
basic training. 


--------------------
\2 According to an official from the U.S.  Army Recruiting Command,
retirement points may be used to calculate retirement eligibility if
the soldier chooses future service in the reserve component.  If the
soldier chooses active service, these retirement points would not be
included in the calculation of retirement eligibility for the regular
component. 


      RECRUITS FAIL TO MEET
      PERFORMANCE STANDARDS
      BECAUSE THEY LACK MOTIVATION
-------------------------------------------------------- Chapter 3:2.2

Another principal reason that recruits are discharged for failure to
meet minimum performance standards is that they lack motivation. 
During our review, we found that all four services have taken steps
to improve recruits' motivation by changing the basic training
environment.  For example, the services have all established remedial
physical training programs for recruits who are not able to meet
physical standards.  The services have also established motivational
and rehabilitation units for those with motivational problems and
injuries.  The services are also concerned with ensuring that drill
instructors do not abuse their recruits.  The Air Force, in
particular, believes that it has made significant improvements in
this area.  A senior Air Force official wrote in June 1995, "The
negative, profane and perhaps even abusive drill sergeant is all but
gone.  [In 1992], almost 10% of our trainees complained of verbal
abuse or profanity.  Today it's 4.1%, and this continues to come
down.  We will get it to zero."

While all four services have similar prohibitions on drill
instructors' treatment of basic trainees, many recruits told us that
some drill instructors motivated recruits through negative
leadership.  The official policy in each service is to treat recruits
with respect.  Navy instructions, for example, state that "the
training process must at all times reflect respect for dignity and
rights of the individual and provide a training environment which is
free from all forms of abuse." Specifically, the instructions say
that "the use of vulgar, obscene, profane, sexually-oriented,
humiliating, or racially/ethnically-slanted language to address or
refer to a trainee(s) directly or indirectly is prohibited.  A
trainee will be addressed only by his/her last name, rank/rate, or by
the word `recruit.'"

Despite this official policy, about one-third of the 126 recruits we
interviewed told us that they were subjected to humiliating treatment
and that this treatment contributed to their desire to leave the
military.  We were told that drill instructors use obscene language
frequently.  What we heard from recruits reinforced findings of Army,
Air Force, and Rand studies that concluded that negative motivation
has a detrimental effect on recruits' desire to stay in the
military.\3 We do not conclude from the anecdotes that recruits
related to us that negative treatment of recruits is widespread
because we cannot generalize from our 126 interviews. 

One study completed by the Army Training and Doctrine Command
(TRADOC) in 1984 found that the way drill sergeants interacted with
their trainees influenced their units' early discharge rates. 
According to the TRADOC report, "Those drill sergeants who outwardly
demonstrated concern for the trainees' well-being, personally gave
additional informal instruction to marginal performers, and who were
effective counselors tended to have more cohesive platoons with lower
attrition." The TRADOC report recommended that training for drill
instructors be bolstered to stress positive leadership and
performance counseling.  TRADOC hoped to create a "win-win" climate
in which they expected higher cohesion and lower attrition.  TRADOC
called this positive leadership philosophy "Insist/Assist." The Army
intended to continue to insist that military standards be met and to
assist all recruits to meet these standards.  TRADOC emphasized that
"The leadership approach of all trainers must be based on positive
leadership techniques and the understanding that trainee performance
is nearly always a function of cadre leadership."

In 1988, a Rand study noted that the Army's basic training attrition
rate was sharply reduced for the fiscal year 1984 and 1985 cohorts. 
Rand concluded that this reduction in attrition "reflects at least in
part the explicit or implicit effect of the Army [1984 TRADOC]
study." That is, Rand believed that this drop in attrition
demonstrated that attrition management can yield large benefits. 
Rand qualified this conclusion by stating that a study would have to
be conducted to ensure that the Army's lower basic training attrition
would not result in a correspondingly higher attrition rate for
enlistees later in their terms.  For the short term, however, Rand
stated that the Army's change in training philosophy and other
changes made as a result of TRADOC's 1984 study, meant that

     With no adjustment in recruit quality or standards, the new Army
     program resulted in 4 and 6 percent more trained high-quality
     men and women available, respectively, in FY85 than in FY83. 
     These effects are comparable in magnitude with those of
     enlistment incentives such as enlistment bonuses and educational
     benefits. 

During our review, we found that the Army had readopted its 1984
philosophy to "Assist/Insist." In a Department of the Army briefing
on attrition dated July 17, 1996, the Army reasserted its 1984
study's conclusions.  The Army stated that it hoped to change its
training philosophy to "stress a positive, reinforcing,
success-oriented environment." The Army said that it hoped to stress
positive leadership, to tell its training personnel, "Don't tear down
to build.  Instead build to build."

In 1993, an Air Force basic training squadron commander came to
similar conclusions to those of the Army concerning the effects of
negative and positive leadership during basic training.  The Air
Force training commander said that at the time he was a basic
training squadron commander in 1988, training philosophies,
particularly recruit motivation techniques, varied among the eight
Air Force training squadrons.  The squadron commander reported that
"some were notoriously negative (and borderline abusive by using such
methods as threatening language, physical intimidation, and excessive
profanity).  Others were more moderate, and several were quite
positive.  .  .  ." The squadron commander believed that negative
styles of leadership would not instill the qualities, values, and
attitudes that were important to the Air Force.  He noted the
following: 

     While there is the need for "rigid" motivation in [basic
     training] to instill certain military skills (i.e., discipline,
     following orders, etc.), and to organize and control large
     numbers of recruits, total reliance on the "negative" style
     won't produce the objectives we seek.  For instance, an airman
     (or flight) that has been motivated mostly by fear, who
     repeatedly over six weeks has been corrected in a degrading or
     harassing manner, may end up resenting authority rather than
     respecting it.  While we may produce someone who follows orders,
     he/she may not do so willingly.  Further, a total emphasis on
     negative motivation risks seeing individual effort dissipating
     in a more relaxed training environment later on.  We can't
     afford that.  Our focus must be on positive motivation.  This
     does not mean that we "coddle" or "carry" trainees, nor does it
     mean eliminating the firmness that is an essential element of
     the Basic Training program.  It does mean a total emphasis on
     professional behavior and proper role modeling. 

The Air Force officer who made these observations said that in the
years that he was a squadron commander, attrition rates varied widely
from squadron to squadron, depending on leadership techniques.  In
his squadron, where he employed positive motivational techniques, the
attrition rate remained at 5 to 6 percent, below the Air Force's
basic training rate of 7.07 in fiscal year 1988 and 7.87 in fiscal
year 1989.  This Air Force officer attributes differences in
attrition rates among the squadrons to differences in leadership
styles. 


--------------------
\3 "Trainee Discharge Program Study," (Army Training and Doctrine
Command, Dec.  26, 1984); Trends in Attrition of High-Quality
Military Recruits (R-3539-FMP, Aug.  1988); and "Developing
Airmanship:  The Focus of the Air Force Indoctrination Process,"
paper prepared for the Air War College, Air University, Maxwell Air
Force Base, April 1993.  As a disclaimer contained in the Air Force
paper suggests, it does not necessarily reflect the official opinion
of the Air War College or the Department of the Air Force. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 3:3

To reduce the attrition of enlisted personnel during the first 6
months of their terms of enlistment, we recommend that the Secretary
of Defense take the following actions: 

  -- Require (1) all the services to review and revise their
     recruiter incentive systems to strengthen incentives for
     recruiters to thoroughly prescreen persons with medical
     histories and (2) the Marine Corps, the Air Force, and the Army
     to more closely link recruiting quotas to recruits' successful
     completion of basic training.  The services may wish to consider
     such ideas as awarding recruiters partial credit for thoroughly
     screening applicants or using a "floating goal" system. 

  -- Direct the services to require all applicants for enlistment to
     (1) provide the names of their medical insurers and providers
     and (2) sign a release form allowing the services to obtain past
     medical information.  Taking these two steps would provide the
     services with the tools to identify applicants' past medical
     problems and add an incentive for applicants to be forthcoming
     in reporting past medical conditions.  To ensure that applicants
     are made aware that the services do follow up in researching
     medical histories, the services could determine how frequently
     they need to obtain applicants' medical records.  To ensure that
     recruiters do not know which applicants will have their medical
     histories researched and that recruiters are not further
     burdened by paperwork, records should be obtained by someone who
     is not a recruiter.  Medical records obtained through this
     process should be included in the file reviewed by the MEPS
     physician at the applicant's preenlistment physical examination. 

  -- Direct the services to revise their "Applicant Medical
     Prescreening Form" (DD Form 2246) and their "Report of Medical
     History" (SF 93) to ensure that medical questions are specific,
     unambiguous, and tied directly to the types of medical
     separations most common for recruits during basic and follow-on
     training. 

  -- Use DOD's newly proposed database of medical diagnostic codes to
     determine whether adding medical screening tests to the MEPS
     examinations and/or providing more thorough medical examinations
     to selected groups of applicants could cost-effectively reduce
     attrition at basic training. 

  -- Place the responsibility for reviewing medical separation files,
     which currently resides with MEPCOM, with an organization
     completely outside the screening process.  Such a review would
     ensure that no conflict of interest interfered with the
     objective review of which medical conditions should have been
     detected by MEPS physicians. 

  -- Direct all the services to test applicants for drugs at the MEPS
     to prevent the enlistment of those who now test positive for
     drugs upon arrival at basic training. 

  -- Direct the Navy, the Marine Corps, and the Air Force to consider
     adopting a program similar to the Army's new Delayed Entry
     Program, under which recruits are encouraged to participate in
     voluntary physical activities with their recruiters and may be
     granted military retirement points for their participation. 
     Also, direct the services to provide those in the Delayed Entry
     Program with access to military fitness facilities and to
     military medical facilities if they are injured while
     participating in physical activities with their recruiters. 


   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 3:4

In commenting on a draft of this report, DOD provided the following
responses to our recommendations.  (DOD's comments are presented in
their entirety in app.  I.)

DOD partially concurred with our recommendation that the services be
directed to review and revise their recruiter incentive systems and
stated that it will direct the services to review their programs and
make revisions as necessary.  Although DOD also stated that it does
not find it advisable to direct the services to revise their
recruiter quota systems unless careful analyses warrant revisions,
DOD also stated that the Navy's new recruiter incentive system, which
subtracts points from the recruiter's quota when a recruit does not
complete basic training, has resulted in lower attrition.  Revisions
such as the Navy's, to "ensure that recruiters maintain a level of
'ownership' in the process," are exactly what we are recommending in
our report.  We believe that the Navy's new incentive system provides
an added incentive for recruiters to take responsibility for the
success of their recruits in basic training. 

DOD also partially concurred with our recommendation that DOD take
actions to improve applicants' reporting of past medical problems. 
DOD stated that it would direct MEPCOM to obtain from applicants the
names of their medical insurers and providers, along with a medical
records release form.  DOD stated that MEPCOM will monitor changes in
the rates of disclosures for past medical conditions and obtain cost
estimates for tracking and obtaining medical records. 

In response to our recommendation that DOD revise its medical
screening forms, DOD partially concurred.  DOD stated that it will
direct the services to review and revise DD Form 2246 and to review
SF 93, which is not a DOD form. 

DOD concurred with our recommendation to use its newly proposed
database of medical diagnostic codes to determine whether adding
medical screening tests to the MEPS examinations and/or providing
more thorough medical examinations to selected groups of applicants
could cost-effectively reduce attrition at basic training.  DOD
stated that it has already established a panel to address this and
related issues. 

DOD did not concur with our recommendation to place responsibility
for reviewing medical separation files outside MEPCOM.  DOD stated
that MEPCOM is an independent agent, completely separate from the
recruiting services, and has no conflict of interest.  DOD said that
the primary purpose of MEPCOM's reviews is to "improve medical
judgment--to educate physicians with the intent of improving
physicians' ability to make the right call given the context of their
screening exams." We believe that this is an important function. 
However, we continue to believe that MEPCOM has a conflict of
interest in determining whether its own physicians should have
discovered disqualifying medical problems in applicants and whether
its own screening methods could be improved.  We believe that an
entity completely outside the medical screening process would be more
able to objectively make these determinations. 

In response to our recommendation that all services test applicants
for drugs at the MEPS, DOD partially concurred.  Before requiring the
Navy and the Marine Corps to change their drug testing procedures,
DOD will require them to conduct a detailed cost-benefit analysis. 
We agree that a cost-benefit analysis would give DOD a sound basis
for this change in policy.  However, we believe that any such cost
analysis should include the cost of transporting, feeding, clothing,
and paying recruits who are now entering the Navy and the Marine
Corps and almost immediately being separated because they test
positive for drugs upon arrival at basic training.  Of the Navy's
enlisted personnel entering basic training in fiscal year 1994, 1,669
were separated for erroneous entry, drug abuse.  Considering the
$4,700 marginal cost of transporting, feeding, clothing, and paying
each of these recruits while they were being separated, the Navy
could have avoided $7.8 million in costs by testing these recruits
before they entered the service. 

DOD concurred with our recommendation that the Navy, the Marine
Corps, and the Air Force consider adopting a program similar to the
Army's new Delayed Entry Program.  However, DOD stated that the
benefits of this new program have not yet been confirmed and that it
is not likely that military retirement points will be included in
this program.  DOD did not state why military retirement points may
not be included. 




(See figure in printed edition.)Appendix I
COMMENTS FROM THE DEPARTMENT OF
DEFENSE
============================================================ Chapter 3



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)


The following are GAO's comments on the Department of Defense's (DOD)
letter dated December 5, 1996. 

GAO COMMENTS

1.  We agree that the problems DOD is experiencing with enlisted
attrition are not new.  We disagree, however, with DOD's statement
that our report offers no new insights.  DOD's problems with
attrition date back to at least 1965, when we reported on the numbers
of recruits who were entering the service with physically
disqualifying conditions.  DOD's problems with data on reasons for
attrition date back to at least 1980, when we recommended that it
improve its data collection in this area. 

Despite DOD's long-standing history of losing one-third of its
enlisted personnel before the end of their first terms, DOD's
department-wide approach to predicting and screening for attrition
has been limited to two primary criteria:  whether recruits have high
school diplomas and whether they score in the upper half of the Armed
Forces Qualification Test (AFQT).  While it is true that recruits
with high school diplomas and high AFQT scores have lower attrition,
we believe that using these two predictors to manage attrition has
reached its limit.  That is, since 1983, the percentage of DOD
recruits who have high school degrees has been over 90 percent, and
the percentage of recruits who score in the upper half of the AFQT
percentile has remained at 58 percent and above.  Yet attrition has
remained at around 30 percent or higher for this entire period of
time. 

We believe that our approach in this report is new in that we are
urging DOD to recognize the limitations of these two predictors and
concentrate on what we see as the real drivers of high attrition: 
the recruiting and examining processes themselves.  To reduce
enlisted attrition, DOD must make systemic, and even cultural,
changes to its recruiting and screening processes.  We believe that
the primary reason for high attrition is that the services are driven
by their obligation to meet overall recruiting goals and end strength
numbers.  Further, we believe that the changes that DOD has made to
its recruiting and screening processes over the years have been only
incremental and that addressing the problem of first-term attrition
will require innovative changes to the recruiting quota system and
screening process such as those we present in our recommendations. 
We also believe that, though the services may be taking individual
initiatives to reduce attrition, DOD needs to gather service-wide
data on the reasons for attrition and take service-wide measures to
control and manage attrition.  We do not mean to imply that the
services are not concerned or are not taking actions to address
attrition.  In fact, we cite many of these efforts throughout the
report.  We do believe, however, that DOD needs to control and manage
attrition service-wide.  Cultural changes to the recruiting and
examining processes will require DOD's direct involvement and its
application of individual service successes among all the services. 

2.  We could not estimate the cost of implementing the
recommendations we make to improve the screening of recruits because
of DOD's long-standing failure to collect service-wide reasons for
attrition.  For example, because DOD does not now know how many
recruits are separated for a particular medical condition, we could
not calculate the cost-effectiveness of adding various medical tests
at the Military Entrance Processing Stations.  As another example,
because DOD does not have service-wide and consistent information on
how many recruits are separated for drug use, we could not evaluate
the relative effectiveness and costs of the services' different
drug-testing policies.  Until DOD collects such service-wide data, it
will be unable to make policy changes based on sound evaluations of
their cost impact.  We support DOD's recent efforts to capture more
accurate medical diagnoses for those who are separated from the
military.  This action indicates that DOD recognizes its need to
collect more specific data on why personnel separate. 

3.  During our review, various officials told us that when there are
several reasons for separating a recruit, the basic training commands
have an incentive to choose the most "face-saving" reason to record
on a servicemember's official release form.  For example, if a
recruit has both motivational and medical problems, the basic
training unit has a built-in incentive to record the medical problem
as the reason for separation both because the basic training unit
would not have to admit that it failed to motivate the recruit and
because the recruit would not have to admit that he or she failed. 
We make this point simply to illustrate the fact that DOD's only
service-wide database on reasons for attrition does not always
accurately capture the actual reason that a recruit fails to complete
his or her first term of enlistment.  Even if military units are not
intentionally miscoding separations, our point remains.  The
separation codes now being captured by the Defense Manpower Data
Center represent DOD's only service-wide information on why recruits
are being separated, and this information is not useful in
determining why attrition is occurring.  We do not mean to suggest
that DOD must use its database of separation program designators for
this purpose.  What we do recommend is that DOD issue implementing
guidance for DOD's separation codes to provide a reliable database
for managing attrition.  DOD could adapt its current database of
separation codes or create a new one to serve this purpose. 

4.  Our recommendations are intended to prevent unqualified recruits
from ever entering the services, not to allow unqualified recruits to
be retained in training and later separated.  We do not believe that
unqualified recruits should be retained.  Rather, we believe that
better screening would prevent unqualified recruits from ever
enlisting. 

5.  The Attrition Advisory Group may in the future be successful in
proposing ways to reduce attrition.  However, at the time of our
review, the Army had not provided us with information on whether the
Group had implemented any initiatives. 

6.  We discuss the efforts of the Office of the Assistant Secretary
of Defense (Health Affairs) to create a medical database on
separations in chapters 2 and 3. 

7.  We state in our report that we believe that the recruiters' award
systems need to be more closely tied to their recruits' successful
completion of basic training.  We believe that the Navy's newly
implemented recruiter incentive system provides such a link, though
it is too soon to predict the new system's effect on attrition.  We
have just begun a review of the recruiter incentive system and plan
to issue a report providing more detail on this issue next year. 

8.  We state in our report that our interviews with 126 separating
recruits do not represent a statistically significant sample. 
Nevertheless, we present the information because we believe that what
these recruits had to say about why they were leaving is valid in
presenting the emotional effects on recruits of being separated from
the military and that any positive changes the services make to
improve the atmosphere for recruits at basic training will enhance
the morale of all recruits.  On the basis of DOD's comments on our
draft report, we have deleted all conclusionary statements on this
issue. 

9.  Our comment on the inadequacy of DOD's data was intended to point
out that the data is not specific enough to allow DOD to determine
whether counseling and rehabilitation efforts would have been called
for in certain instances.  For example, if recruits who have
disqualifying medical problems are being discharged under a
separation code that indicates performance problems, then counseling
and rehabilitation for these recruits would not have any effect.  As
another example, because recruits with motivational problems can be
discharged under separation codes that indicate medical problems, DOD
cannot determine whether motivational techniques might have worked to
help these recruits. 

10.  We discuss the Air Force's efforts to reduce attrition, in a
general way, in chapter 3. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix II

NATIONAL SECURITY AND
INTERNATIONAL AFFAIRS DIVISION,
WASHINGTON, D.C. 

Sharon Cekala
Elliott Smith
Beverly Schladt
David Moser
MaeWanda Jackson
Jai Lee
Charles Perdue
Bill Beusse
Waverly Sykes
Ernie Jackson
Nancy Ragsdale


*** End of document. ***