Defense Inventory: Defense Logistics Agency Needs to Better Manage Procurement Leadtimes

Published by the Government Accountability Office on 1990-05-02.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                                     United   States   General   Accounting   Office
                                     Report to Congressional Requesters                .

      ! May 1990
                                     Defense Logistics
                                     Agency Needs to
                                     Better Manage

                              :   Im3TmcTED     --Not    to be reieased outsic.b the
                                  General Acmnnting Office unless tqxciflw
                                  approved by the Off’ice of Congtwdmml


*.    .   GAiO/NSIAD-90-124
                   United States

GAO                General Accounting  Office
                   Washington, D.C. 20548

                   National Security and
                   International Affairs Division


                   May 2,199O

                   The Honorable John Glenn
                   Chairman, Committee on
                      Governmental Affairs
                   IJnited States Senate

                   The Honorable Earl Hut to
                   Chairman, Subcommittee on Readiness
                      Committee on Armed Services
                   IIouse of Representati\,os

                   As you requested, we evaluated the Defense Logistics Agency’s (DLA)
                   controls over procurement leadtimes used in determining when to initi-
                   ate a buy. You specifically wanted to know if DLA has adequate manage-
                   ment controls to ensure leadtime estimates being used are appropriate to
                   support mission requirements. If they are not,, significant adverse
                   cffrcts can occur.

                   Overstated leadtimes can cause increased investment for larger invento-
                   ries, greater chances of buying excess material, and increased termina-
                   tion costs if requirements change. Irnderstated leadtimes can cause
                   shortages of needed supplies, which could affect the operational readi-
                   ness of weapon systems or their components.

                   As of September 30, 1989, average procurement leadtimes at DIA supply
                   ccntcrs ranged from 4 months to 17 months. This required DLA to main-
                   tain on-hand and on-order inventory levels valued at about $5.4 billion.
                   The objectives of our review were to evaluate (1) the DLA practices and
                   procedures for controlling procurement leadtime estimates and (2) DLA’S
                   efforts to reduce actual leadtimes. To do this we identified and tested
                   controls at two of ILA’S six supply centers-the    Defense Electronics
                   Supply Center (L)M(‘), Dayton, Ohio, and the medical supply activities of
                   the Defense Personnel Support Center (a~), Philadelphia,
                   Pennsylvania. At these support centers, we reviewed the reasonableness
                   of the leadt,imes used for a small (nonprojectable) random sample of
                   stock items. (See appendix I for a more detailed description of our objec-
                   t ivcs, scope, and methodology.)

                       has not implemented adequate controls to manage and minimize pro-
Results in Brief   DIG
                   curement leadtimes as directed by the Department of Defense (non). Our
                   sample items at the t M’() supply renters had leadtimcs that were either

                   Page 1                                      GAO/NSIAIHO-124   Defense   Invrnlory

                       for production leadtim<>. The longer the procurement leadtime, the more
                       on-hand and on-order stock is needed t,o meet customer demands and
                       ensure a continuous s111)ply of stock items.

                       Longer than necessary Icadt.imes cause DLA to purchase items sooner
                       than needed. This results in an earlier commitment, and obligation of
                       funds and increases the risks of buying too much or too little stock. For
                       example, if demand for an item decreases during a long leadtime period,
                       the item being purchased c.an exceed the services’ needs. This would
                       require DIA to cit,htlr c~anc4 the item on order, dispose of the item, or
                       store it for longer periods. If, however, demand increases, longer lcad-
                       times slow MA reac*tion time to get more items in stock, increasing the
                       risk of being out of stock

                       Overstated leadtimes could also increase safety level stocks. These are
                       st.ock levels wi~ich ark ant horized to be on hand as a cushion should
                       demand unexpectedly increase or the actual procurement leadtime take
                       longer than forecasted. Safety level requirements are based on a com-
                       plc~ formula and c~hangt~sin leadtime have a direct effect on the safety
                       level requirements of i~tdividual items. DI,4 officials estimated that a lo-
                       day change in leadtim<\ wollld have a corresponding change in safety
                       1~~1 rc~quircmcnt,s (11’I .7 days. Although WC do not know what effect
                       this would have at th(, IX.:\ level, I)OL)estimates that for each day lead-
                       times are reduced. MI l-wide. $10 million is saved in smaller quantities of
                       safety level stocks.

                       DIA  manages about tiX5,OOO consumable supply items that, are based on
                       a forecasted lcadtimt~ 11sc~1  to determine inventory requirements and
                       when to initiate a t)ll> (‘onsumable items are discarded after USCas
                       opposed to being r(~pmr~i or serviced. They include such items as cloth-
                       ing. food. medical, industrial. construction, and electronics supplies.
                       About 75 percent of t hc items managed by DLA have a unit price of less
                       than $50.

                       On December 9. 1985, IIOI) issued an instruction’     that required all ser-
DLA Lacks              vic,cs and ILA to cstabhsh specific management        controls to minimize pro-
Appropriate Leadtime   curement leadtimes. Xlt hough LLA headquarters          and the supply centers
Controls               WY visited had taken s\om(l actions to implement       this instruction,
                       imgro\-rments are still ncbc,dcd.More specifically,      we found that DIA
                       and/or the centers Ilild no1

                       Paye 3                                         GAO:NSIAD-90.124   Defense Inventory
                               contract award. Tht> Tt~c~hnical Directorate gets involved in the procure-
                               ment action, primarily to identify product, vendors and ensure that
                               items meet specificat ions. Hased on our discussions with Supply Direc-
                               torate officials, thr supply tnanagcment review, if done! would include
                               only the procedures and actions of the supply directorat,e.

DLA Has     Not Set            The DOUdircctivo rclquircs IKA to establish procedures to ensure that
                               leadtime estimates art‘ based on representative procurements-those
Standards    or Developed
                               with routine and rc,c,urring procurement actions and circumstances. To
Accurate    Data to Identify   do this, the directivca rc,quires IL+. to develop standard times (in days)
Abnormal     Leadtimes         for carrying out, t,he different phases of routine procurement actions.
                               ‘L‘hcse standards allow IHA to identify abnormal delays or nonroutine
                               cvcnts and keep them from inflating future leadtime estimates. The
                               directive also requires IL.\ to maintain historical information on all
                               procurements (act,l~al tmlc’s for various phases of the process) and use
                               rcprcscntative times from t,his database to establish forecasted lead-
                               t,imes. This requirctrtc~nt is dt++cd    to help ensure that the forecasted
                               Icadtime used by IL \ 15appropriate and will not unnecessarily distort
                               inventory requirtmon15.

                               At the two locations wt’ visited--r)rwc and r)Esc-we found that these
                               rcquirc,ments had not t)c,cti implemented. Moreover, neither organization
                               had instituted effect ivtb alWrnative procedures to ensure the forecasted
                               leadtimes rcpresentcd rout inc or recurring procurement actions. In fact,
                               our tests of 57 supply items. selected from a universe of 24,000 high
                               demand value items, shoived that 37 (67 percent) had administrative
                               Iradtimes that includrd cvt~nts we believe to be nonrepresentative.
                               Thrst~ events inclutic~rl lost purchase requests, nonrecurring time to iden-
                               i ify a nc’w source. man~~gt~tncnt direct,ed delays, contractor requested
                               delays. delays causc~i tly training and excessive buyer work load, and
                               onct-time technical rc\ ithws.

                               LNXSC buys electrical cotit acts for the three military  services. Forecasted
                               adtninistrative   lcadt,imc, for the most recent, purchase at the time of our
                               rcavicw was 357 days. To dctcrmine this number, DLA used a weighted
                               average formula whic+l gives greatest, weight to the two most recent
                               purchases, assuming t ltt’y will be more representative of the future.
                               Therefore. in assessing the accuracy of the 357 days. WClimited the
                           l   It took 150 days longer than normal to process the buy in part because
                               interns, being used to process the buy, were in school for 30 days. DPSC
                               officials were unaware of the causes for the remaining 120-day delay,
                               and records were insufficient to show the reasons.

                               If I)I’SC had eliminated the times associated with these nonrecurring
                               events, WC estimate (using DLA’S weighted average formula) that admin-
                               ist rative leadtime for the most recent buy would have been reduced by
                               i 1 daysPfrom       177 to 1Oti. This would have reduced inventory and
                               safety level requirements by $990 and $167. respect,ively.

                               For all 57 items we revrewed. inventory requirements and safety levels
                               could have been reductd by about $2 12,000, if DLA had a system to iden-
                               tify and eliminate unusual or nonrecurring events from forecasted lead-
                               times. We rannot project these findings to DLA’S total inventory based on
                               ow    small sample. Ilowc~vcr, our review indicates that significant savings
                               may be possible considering that DU manages about 686,000 inventory
                               it,ctns that use leadtimrb to determine requirements.

                               For the above cxamplcs, WC’were able to identify the causes of signifi-
                               cant delays only after tlxtensive reviews of the contract files and talking
                               wit,h the appropriate buyers and item managers. For many of our sam-
                               ple items, we identified delays but were unable to determine a cause.
                               This was due to insufficient information in contract files and buyers and
                               item managers not being able to recall contract events. In some cases,
                               buyers had retired, transferred within LKA, or quit. The lack of records
                               prevented reconstruct ion of the causes of procurement delays. DLA offi-
                               cials believe they art’ doing the best, they can with available resources to
                               carry out the 19% directive. We recognize that the current DL\ proce-
                               dure. which is intended to identify gross differences between projected
                               and actual lcadtimcs, has helped identify some inflated leadtimes. IIow-
                               c\‘cr. the procedure dtlpends extensively on the individual initiative and
                               jrtdgmcnt of itctn managers. During our review we found item managers
                               tried to carry out tlrtx procedure but usually lacked sufficient informa-
                                tiott to identify abnormal leadtime events.

Use of Contract Delivery       LLA  computes foreuastt,d production leadtime twice on every purchase.
                               Dcpcnding on the st agr’ of the purchase, the forecast is based on a
Dates Distort Forecasted
                               weight t,d average of tit her the current contractual or the actual delivery
Leadtimes                      time and the old avtsragr in the automated system. At the time DIA
                               issues II contract, it ust~ the contract delivery date to update forecasted
                               Icadtimes in the automated system for t,he next buy. Actual delivery

                               t%gt.   7                                     GAO/‘NSIAD-W-124   Defense   Inventory

completed in March 1987, centered on ways to shorten administrative
leadtime by increasing automation and improving contracting t WII-
niques and management practices. The plan did not inc4udr ways to
reduce production leadtime even though it now accounts for about CO
percent of total procurement leadtimcl.

The IJnder Secretary’s December 1986 memorandum enc,ouragf>d the
reduction of production Icadtimc. The memorandum includrd rc’c.om-
mendations to reduce procurement leadtimc that had btlcsn mad<’ by the
Logistics Management Institute in its 1986 report on pro~ut’cmcnt lead-
times.” However, WV found that DESK and I)PSC‘ had not implcmcnt~~d spc-
cific recommt>ndations for reducing production Icadt imc by having
supply units request thcl cont,ractors’ best delivery times. Thc~ timt,s
can be used, where possible. to rcduc*r forecasted leadtimes for stihsC-
quent buys and in selecting a contractor. In fact. the’ buyers at theso
centers normally rcqnc‘st contractors to deliver t,hr items on the dat c
specified in ULA’S automated inventory system. They do not ncc-t,ssarily
want the contractor’s btxst delivery times on all pnrctrasc~s.

I)ESC~ officials explain4  that a contractor’s best delivery time, should be
reserved for items tllat ark Ilrgently needed because of supply
shortages. In their opinion. contract,ing for the best delivery time in all
cases would cause vcsndors to inflate future delivery times to avoid con-
tract penalties for late, deliveries. In turn, they said that pressure to
achieve the best delivc>ry time on all buys would rcducc thtl ~cndo~~‘s
willingness to providt, quick deliveries on urgent tx~ys.

If UIA requested thcl b& delivery time in compctitivtl solicitations. con-
tractors that inflat ~1 d<,livtlry times to avoid fut urc pcnalt ies would
lessen their chances of’ receiving the contract awards. This. in our view.
is enough of an inccWive to obtain some improvemom in dclivcxry t imcs.
In addition, the best dt>livcry time does not necessarily mean “urgt’nt”
delivery time. In ernc~rgc~ncysituations, LLA still has tlrc‘ option 01’ offcr-
ing incentives for the, contractor to meet more> urgc,rrt schedultx

The bottom line, however. is that the uw centers arc’ not folknving the
llnder Secretary’s 11186direction and arc missing an opportunity to
reduce actual produc,tion leadtime. This is importwnt~ hccxllstl IHA’s anto-
mated inventory system uses both the contracted and a(.( ual prodllc,tion
leadtimc to generat (3t’or(~(.astcd lradtimcs for fu(utx> buys.

Paye 9                                         GAO,‘NSIAD-90-124   Ikfiwsr   Invmlory

                          We, therefore, recommend that the Secretary of Defense direct the
                          Director, DLA, to

                      . conduct periodic, objective, and comprehensive management reviews at
                        the supply centers to assure all center directorates involved with lead-
                        time comply with applicable DOD procurement leadtime policies and
                      . set sound standard time frames for key administrative     leadtime events,
                        develop reliable information on the actual time taken for such events,
                        and then compare standards to actual times to remove unusual events in
                        forecasting leadtimes;
                      . use only actual delivery dates to forecast procurement leadtime unless
                        they are not available or considered unrealistic; and
                      . request a contractor’s best delivery time to reduce forecasted leadtime
                         for future buys and where possible give that delivery time some weight
                         in contract awards.

                          In its oral comments on our draft report, DOD generally agreed with the
Agency Comments and       report. DOD acknowledged that greater emphasis needed to be placed on
Our Evaluations           both accurately forecasting and reducing procurement leadtimes. DOD
                          stated that DLA would review its supply centers’ (1) procedures and ini-
                          tiatives to minimize and accurately forecast leadtimes, (2) criteria and
                          practices for identifying nonrepresentative  buy situations with a goal of
                          developing consistent agencywide guidelines by April 1991, and (3)
                          parameters and related procedures for reviewing leadtime updates.

                          DOD    commented that aggressive action in this area will implement com-
                           mitments made by the Secretary of Defense and the lJnder Secretary of
                           Defense (Acquisitions)  to pursue a cultural change within DOD'S acquisi-
                           t ion process in which cost-effectiveness is given increased priority.

                           1101)acknowledged that ILA had not conducted the required periodic on-
                           site supply management reviews since 1986. In DLA's opinion, the
                           payback in terms of problems identified and corrected did not justify
                           the expenditure of the resource required. Rather than doing the
                           required periodic revickvs, ILA performed “intensive”    reviews on spe-
                           cific known problem areas. nou also commented that since 1986, DLA hat
                           rontinucd to perform its contract management reviews, which concen-
                           t rated on contracting practices and included administrative   leadtime
                           issues. Thus, DOD belicvcts ILA is in compliance with the DOD instruction.
GAO, NSIAD-90.124   Defense Inventory
Appendix   II

Major Contributions to This Report

                        Richard A. Helmer, Assistant    Director
National Security and
International Affairs
Washington, D.C.

                        Richard    Strittmatter.   Assignment Manager
Cincinnati Regonal      Michael    J. Hazard, ISvalrlator-in-Charge
Office                  Donald    I,. Allgyer, Site Senior
                        Katrina    D. Stewart, I:valuator

                        IIarry IL Benchuff. .Jr., Assignment   Manager
Philadelphia Regional   -John R. Kirstein, Sittl Svniol
Office                  George A. Scott, Evaluator

(391613)                Page 15                                         GAO/NSIAD-90-124   Defense   Inventory
    Requests for copies of GAO reports     should be sent to:

    U.S. General Accounting    Office
    Post Office Box 6015
    Gaithersburg, Maryland     20877

    Telephone   202-275-6241

.   The first five copies of each report   are free. Additional   topic
    $2.00 each.                                                           I

    There is a 25% discount    on orders for 100 or more copies m;
    single address.                                                       I
Appendix   I

Objectives, Scopeand Meihodology

               The Chairmen, House Committee on Armed Services and the Senate
               Committee on Governmental Affairs, asked us to evaluate the Defense
               Logistics Agency’s (DLA) procurement leadtime practices and proce-
               dures. The underlying concern was that leadtimes may be too high. Our
               objectives were to evaluate the practices and procedures for controlling
               DLA leadtime estimates used for determining    inventory requirements and
               I)LA’S efforts to reduce actual leadtimes. We performed our work at the
               Office of Secretary of Defense and Headquarters, DLA, Washington, D.C.;
               Defense Electronics Supply Center, Dayton, Ohio; Defense Personnel
               Support Center, Philadelphia, Pennsylvania; and the Logistics Manage-
               ment Institute, Washington, D.C.

               We reviewed pertinent documents and reports and met with officials to
               determine how the DLA supply and procurement systems worked. We
               identified key DLA systems and controls to manage leadtimes and tested
               their effectiveness. We used DIA computer programs, reports, and

               We used a case study approach to determine if leadtimes were too high
               and if controls were in place and working. An initial random selection of
               50 high demand value replenishment items was drawn from D1.A supply
               computer tapes, for both the Defense Electronics Supply Center and
               Defense Personnel Supply Center, as of September 1988. Our review of
               controls focused primarily on administrative  leadtime since DLA has
               greater control over actions that make up administrative   leadtime. We
               completed evaluations on 34 electronics items and 23 medical items.

               We were unable to project the results of our sample because of its small
               size and the variance of the results. Nevertheless, our analysis does indi-
               cate that adequate controls are not in place and administrative  leadtime
               has not been minimized as required by thr Department of Defense. We
               conducted our review between August 1988 and November 1989 in
               accordance with generally accepted government auditing standards.

               Page 14                                      GAO/NSIAD-90.124   Defense   Invrntory

We do not take issue with the statement that L)I.~\is doing “intcnsivc”
reviews to correct major problems at certain supply centers. IIowt~vcr.
since the periodic reviews were curtailed and our rcvicw has indicated
that the instruction’s objectives (i.e., ensuring that lcadtimes were mini-
mized, policies wtbrc being followed, and Itadtimr forecasts werr acc~
rate) have not ber>n achieved, we do not bt%evc r)I,z\ is in full compliance
with the r)ot) instrllct ion.

As arranged with your offices, unless you publicly announce its contents
r,arlicr. we plan no further distribution of this report until 30 days from
its issue date. At that time, we will send copies to the Chairmen, Sltb-
committees on Dtlfcnsc. Senate and Ilouse Committees on Appropria-
t ions; Senate and LIouse Committees on Armed Scrviccxs; t hc Director,
r)l& the Director. Office of Management, and Budget; and other intcr-
clstctl parties. WC will make copies available IO othclrs upon request.

If you or your staff have any questions, please call me on 2’6841%
Major contributors to thrl report are listed in appendix II.

Donna M. lleivilin
Director, Logist ic.5 Issncs

Page 12

                  Part of the reported 12.percent reduction in leadtime resulted from
                  across-the-board changes in leadtimes made by the two Centers we vis-
                  ited. These changes were not, however, based upon actual changes in
                  procurement processing time. At DESC cuts in leadtime were made to
                  reduce requirements and help solve a severe funding problem. While the
                  funding problem may have been mitigated, the cuts distorted DKT'S
                  leadtime data. DIW officials admitted these across-the-board cuts were
                  arbitrary and may have produced negative effects on supply operations
                  had it not been for decreasing demands at the time. r)r’sc also made
                  across-the-board cuts because its leadtimes were too high. Moreover. DI,.~
                  officials informed us that the Centers all made across-the-board cuts
                  because leadtimes were too high DLA-wide.

                  In its oral comments on our draft report, DODsaid that it agrees that DLA
                  needs to do more to reduce leadtimes. As an example, DOD said that DLA
                  is developing automated systems to use “best delivery” data on future
                  buys. Its Contract Decision/Analysis    Support Program initiatives are
                  designed to provide this capability. According to DOD, t,arget dates for
                  implementation of’ this initiative extend beyond 1995.

                  We believe that conditions found at the two DLA centers indicate a need
Conclusions and   to establish better controls and procedures to manage and minimize
Recommendations   leadtimes as required by the DOD policy directive. Specifically, r&i offi-
                  cials had not (1) periodically reviewed leadtime practices at these sup-
                  ply centers to ensure that policy was fully implemented, (2) established
                  standards for leadt imes or collected complete and accurate informat,ion
                  on actual leadtime, so that proper evaluations could be made to elimi-
                  nate nonrepresentative      procuremtnt actions from forecasted leadtime.
                  or (3) used accurat~e contract delivery dates in computing forecasted
                  lcadtimes. While r)I.i\ officials did take some steps to rcducc administra-
                  tive leadtime. they virtually ignored production leadtime.

                  The primary emphasis of DLA officials has been on ensuring that stock is
                  available to meet customer needs. Since using longer lcadtimes can
                  increase stock availability,  there is little incentive to reduce leadtimes.
                  We believe it is import ant that DLA officials implement effective manage-
                  ment controls over procurement leadtimes because r)td\ has large invest-
                  ments in supplies to srlpport leadtime requirements.

                  Page10                                        GAO/NSIAI%YO-124Defensr   Inventmy

                       time is used to updat,c the forecast only after 51 percent of the largest,
                       contract line item has been delivered. However, if actual delivery times
                       differ from the contract times, t,he forecasted leadtimes can be distorted.
                       R7e compared actual and contract delivery times on 109 cont,racts for
                       the 57 supply items WC reviewed at IIESCand L)t’s:(’to det,ermine if these
                       dates differed and, if so, the extent of the difference. We found only one
                       case where the contract and actual delivery time was the same. For the
                       other 108 contracts:

                     - Actual deliveries for 58 occurred on the average of 51 days after the
                       rontract delivery date. This understated inventory requirements and
                       could have resulted in shortages of items valued at about $200.000.
                     - Actual deliveries for 50 occurred on the average of 6 1 days before the
                       contract delivery date. This overstated inventory requirements and
                       could have resulted in unnecessary items vahled at about $99,000.

                       According to KID, the reason DLA uses estimated contract delivery dates
                       as much as it does to forecast leadtimes is linked to DLA'S inability in the
                       past, to react to rapidi>. increasing production leadtime by suppliers.

                       The DOD leadtime instruction clearly states that if information to be used
                       to forecast leadtimes is known to be inaccurate, it should not be used.
                       TJsing inaccurate estimated delivery dates, even if from a real world buy
                       situation, creates inaccurate leadtime forecasts and does not solve the
                       problem as originally intended. The inaccurate delivery information
                       becomes the major factor in forecasting leadtime for the next purchase
                       of an item, because it is the basis for computing that purchase’s esti-
                       mated contract delivery date. Therefore, DLA'S current system for corn-
                       puting lcadtime continues to be based on inaccurate information and
                       perpet,uates inaccurate leadtimes. That is why we believe that ILL needs
                       to curb its practice of using estimated contract delivery dates in fore-
                       casting production lewdtimes.

                       Although DLA has reduced leadtimes by 12 percent in response to the
DLA Can Do More to     1986 directive, we found that ILA had not implemented t)or) recommen-
Reduce Leadtimes       dations with regard to production leadtime and that the 1P-percent
                       reduction resulted in part from funding limitations and across-the-board

                       In December 1986, the IJnder Secretary of Defense (Acquisition and
                       Logistics) directed 1~4 to develop a specific plan to significantly reduce
                       procurement leadtimes by the end of fiscal year 1988. The DLA plan,

                       Page8                                         ~AO/NSIAD-90-124Defe~~srl~1vrntory
     sc*opc of our evaluation to these two most recent purchases. Our evalua-
     tion identified several nonroutine or nonrecurring events that p1.A
     should have excluded in arriving at forecasted leadtimc.

. The buyer had a high work load of 500 purchase requests at the time
  tlw system rcconuncwlcd       t hc purchase. As a result, 91 days lapsed
  before the buyer ~uld start the procurement process. Normally this
  takes about 5 days. During this delay a second buy was initiated and
  combined with the first buy, increasing the size of the buy to over the
  S25.000-ceiling for small purchases. This required even more steps t,o be
  taken and additional proc’urcment delays.
. A potential contractor rcyucsted and received a X)-day extension on the
  time to submit, a bid. This did not occur on the previous buy and should
  not happen on every pllrchasc.
l Two bidders included unapproved parts in their proposals. The Techni-
  WI Directorate’s rc>vic,w and approval of these parts added 100 days to
  the procurement proc~s. On the previous contract, no technical reviews
   were required and no alWnate parts were bid.
l l’hc contractor refustsd to submit cost and pricing data, as required for
   purchases over d 100.000 claiming an exception to the regulations. The
   resolution of this r~nu~al t)roblem took r%%C:   89 extra days. Only I out
  of 1 1 contracts for tnis ;t.tm since IR81. has exceeded $lOO,OOU.

      If DESChad climinatc4 I hc t imcs associated with these nonrecurring
      events. we estimate> (u’iing DIA’S weighted average formula’) that admin-
      istrative leadtimc I’or t IW most recent buy would have been reduced by
       130 days-from    357 t 1) 227. We estimate that this would have rcduc*ed
      kadtimc and safety IPWI requirements by $8,230 and $1,400,

      oInc‘-Medical buys cannula sets (tubes to drain fluid from the body) for
      the three military sc~~~icr~s.E’orccastcd administrative leadtime for the
      most, recent purchase at the time of our review was 177 days. Our evalu-
      ation of the two most rcac.txntpurchases showed the following nonrout,inc
      or nonrecurring events that r)l-&.should have excluded in arriving at
      forcacasted lt‘adtimtt

    . A 1987 reduction in force eliminated some buyers at WW, and it took an
      extra 5fi days to proccn\s the purchase rcquclst.

      Page Ii                                     GAO,'NSlAtWO-124   Drfensr   Inventory

                        . conducted periodic supply management reviews,
                        * established leadtime standards or developed complete and accurate
                          leadtime information, and
                        * used appropriat,c actllal delivery times as a basis for estimating procure-
                          ment leadtime.

                          DOD  recognizes that procurement leadtimes need to be controlled. Its pol-
                          icy states that if lcadtimcs are permitted to increase unchecked. stock
                          quantities can also rise, increasing the cshance of buying items in excess
                          of needs.

DLA Has Not Conducted     The 1985 r)or) inst ruc.tion requires DLA to conduct periodic supply man-
                          agemcnt reviews to onsur( leadtimes arc minimized, policy is being fol-
Supply Management         lowed, and leadtime forecast.s arc accurate. According to I)I,A
Reviews                   Kcquirements Ilranch officials in the Supply Directorate (rcsponsiblr for
                          conducting the revi<‘\\ ). these supply management reviews have not
                          been conducted for WUWI years. They could not recall when the last
                          one was done, nor 1)I’(bvidc ;I report documenting that a review had been
                          done in the past. ‘l’h~~ Hranch officials recognized the importance of
                          the reviews but said i h(~r Hrancb lacked funds to do them.

                          As a substitute for t INSrc>views. the officials said that leadtimes were
                          discussed at annual s;11pply conferences attended by supply managers
                          l’rom IL\ htadquartctrs and supply centers. They also told us that, lead
                          times were discusscltl during management visits to the supply centers.
                          One official said the l3rxnc+ does not maintain records to support the
                          frequency or conttsut in Lvhich leadt,imes were discussed.

                          The Supply Directorat c has litt,le incentive to minimize leadt,imcs. The
                          focus of supply managt’mc~nt is primarily on having stock available to
                          meet customer demands. Inflated leadtimes increase on-order and on-
                          hand invent,oritks ~vlric II improves t,he directorate’s ability to meet cus-
                          tomer needs and ~sllrt~s management indicators of supply availability
                          rtbmain high. Grtlattsr c,mphasis on supply availability     overshadows r)lA’s
                          mission rcquircmtW 10 c,ost-r~ft’c~ctively support the military services.

                          Although responsibilil)’       for conducting the supply management reviews
                          lies within r)I,A’s Srlpply I)ircctorat~. two other directorates-the     Con-
                          tracting and I’rodrlc~tion Directorate and the Technical Directorate-also
                          have w major role in managing Icadtime. The Contracting Directorate
                          incurs most of the ;I(Iilllrlistrativc‘  lcadtimc days in preparing for thtl

                           Pay? 4                                        GAO/NSIAtS90-124   Defense Inventory

               overstated or undcrst atcd, thus increasing the risk of buying too much
               or too little stock. Spccific*;rlly, 1~4 has not

             . conducted required sr~pply management reviews to ensure that lead-
               times are accurately forecast cd and management actions are being taken
               t,o minimize leadtimf,.
             . set standards for the various stages of the buying process or developed
               complete and accurate information so that, nonrepresentative    procurc-
               ment actions can be itlent ifiod and eliminated from the database used to
               forecast procurement lcadt imes; 01
             . used realistic deliver>, date‘s to forecast leadtimes.

               Although LL+ has taken m(%sures to reduce the time it takes to award
               contracts, it has not tried to reduce production and delivery times by
               obtaining the best dclivc>ry dates from contractors. Production and
               delivery tiIncs acc’ount for fi0 pcrccnt of total procurement leadtime.

               In late 1985, non issrwti a policy directive that identified management
Background     controls to minimix> 1~roc~rIremcnt leadtimcs. In December 1986, I)OL)
               directed DLA to dcvr>lop a sptcit’i(, plan to reduce leadtimes by shortening
               the amount of tinIt> nc~cd~~dto order and deliver supplies. A 1989 I)LA
               lcadtime indicators rc~por-t shows that the forecasted lcadtime (measured
               in averagc dollar wrxighttd kadtime days) had been reduced by about 1
               month, or 12 pcrccsnt sinc,c 1CM-from        323 to 286 days. I During this
               period leadtimc rcqIIirc>Inc>nts went from about $6.0 billion to $5.4

               The 1~4 supply ccntc~ use an automated inventory system that periodi-
               cally compares in\,clnt ory requirements with assets on hand and on
               order and recomm(Ms ;IC,~ions to inrreasc or reduce stock levels to meet
               projected demands. OIW of the important considerations in the auto-
               mated system is t hcl f’orocxstt~d pro~uremcnt, lcadtime. This is the
               amount of time IN. 1 csstiInatc,s that it will need to buy items in advance,
               so it will not run 1IlIt cbl‘sr~pply before the iteIns are delivered.

                The forecasted proc~ rcimcnt lcadt imc is divided into administrative
                lradtime (the time Ix~clIlirr~d to initiate a buy and award a contact) and
                production leadt iIrrc\ / t hlb t imc for the contractor to produce and deliver
                the item). ~~1’s 1989 Ioacltirnr~ data show t,hat about 40 percent of the
                procurement lcadt irucs Is l’or administrative       lcadtime and 60 percent is

                Pa@? 2                                          GAOjNSIAD-90.124   Defense   Inventory