oversight

Navy Supply: Procurement Leadtime Forecasting Needs Improvement

Published by the Government Accountability Office on 1990-05-18.

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

                             ,.




                                                                                   .. _

                                        ”            :
~                  .~,                                              -,                        . .        :
    _.I.   .   .                            :   ‘:       ‘..   _.        ._   ~.          .         :.
                         :        ‘.,
National      Security and
International        Affairs Division

B-238716

May l&l990

The Honorable Les Aspin
Chairman, Committee on
  Armed Services
House of Representatives

Dear Mr. Chairman:

In response to discussions with your office, we reviewed the Navy’s process for determining
procurement leadtime requirements. We found that the Navy can make improvements in
forecasting these requirements.

We are sending copies of this report to the Chairmen, Senate Committee on Governmental
Affairs, House Committee on Government Operations, Senate Committee on Armed Services,
and Senate and House Committees on Appropriations;     the Director, Office of Management
and Budget; and the Secretaries of Defense and the Navy.

This report was prepared under the direction of Martin Ferber, Director, Navy Issues, who
may be reached on (202) 2756504 if you or your staff have any questions. Other major
contributors are listed in appendix II.

Sincerely yours,




Frank C. Conahan
Assistant Comptroller           General
Executive Summq

                                                                                                      ;
                                                                                                      1


Purpose                                                                                               8
                   that the Navy estimated was needed while awaiting additional deliveries            1
                   ranged from $2.1 to $2.9 billion a year. At the request of the House
                   Committee on Armed Services, GAO evaluated whether the Navy was
                   using credible procurement leadtimes in determining the need for these
                   parts.


                   In determining inventory requirements, the Navy considers estimated
Background         procurement leadtimes, safety levels, and economic order quantities.
                   The procurement leadtime level is the amount of inventory needed to
                   meet normal demand during the time required to order and receive addi-
                   tional inventory. A safety level is the amount needed to meet fluctua-
                   tions in demand and leadtime. The procurement leadtime level is further
                   divided into administ,rative leadtime and production leadtime. The
                   amount of material needed to meet demands from the time a need is
                   determined until a contract is awarded to purchase the material is called
                   administrative   leadtime, The amount of material needed to meet
                   demands during the time from the award of a contract until receipt of
                   the material is called production leadtime.

                   Because the Navy’s Aviation Supply Office manages approximately
                   162,000 aviation items, determining valid requirements is a challenge;         1
                   moreover, this magnitude also makes inventory management vulnerable
                   to inefficiency and waste.
                                                                                                  ;
                                                                                                  Y
                   The Aviation Supply Office can improve determinations of procurement          te
Results in Brief   leadtime requirements for aviation parts. Administrative     leadtime         1
                   requirements were not always based on actual experience. GAO found            1
                   that at one point, the Supply Office had arbitrarily increased the admin-     !
                   istrative leadtimes for all items by 9 months. In calculating production      t
                   leadtime requirements, the Supply Office did not consider some actual         1
                   experienced leadtimes even when these leadtimes were more realistic.
                   The Supply Office also did not routinely obtain contractor estimates of       t
                   leadtime or compare them with actual performance.
                                                                                                  I
                   Inaccurate leadtimes have significant adverse effects. Overstated lead-        I
                   times cause added investment for larger stock levels, greater chances of      ij
                   buying excess material, and increased termination costs if requirements       j
                   change. Understated leadtimes cause material shortages and reduced
                   readiness of the units needing the material.                                  1
                                                                                                 t


                   Page2                                   GAO/NSIAD-90-78NavykadtimeForecasta   1
                                                                                                 1
---
                             ExecutiveSummary




                             In a random sample of 21 items, GAO found either overstated           or under-
                             stated requirements on all items. The overstatements totaled          $2.2 mil-
                             lion and the understatements totaled $839,000. With 162,000            items
                             having leadtime requirements of over $2 billion, the potential         for signifi-
                             cant efficiencies and dollar savings is great.



Principal   Findings

Administr -ative Lead time   Although Department of Defense (DOD) policy requires that administra-
                             tive leadtime forecasts be based on historical information that has been
Estimates Are Not            collected for representative procurements, the Aviation Supply Office
Accurate                     groups its 162,000 different stock items into only a few groups and uses
                             the same standard leadtime for all individual items within a group.
                             These standards may reflect the average leadtime experienced for an
                             entire group, but they do not reflect the differences in individual lead-
                             times within a group. The Army and the Air Force estimate administra-
                             tive leadtimes for individual items, and both DOD and Navy officials say
                             it is feasible for the Navy to also do so.

                             GAO'S  analysis of 2,467 items purchased during a recent period showed
                             that actual administrative leadtimes for 863 purchases varied from the
                             standards by at least 6 months. In other words, the Aviation Supply
                             Office either overstated or understated administrative leadtime require-
                             ments by 6 months or more in 35 percent of the purchases analyzed.

                             GAO  also randomly sampled 150 stock items but, because of inaccuracies
                             in the Aviation Supply Office’s files, was able to obtain requirements
                             data on only 21 items. GAO'S review of these 21 items showed that the
                             number of days used in the administrative leadtime requirements com-
                             putations were overstated for 15 items and understated for 6 items. The
                             value of the overstated requirements was $645,000 and the value of the
                             understated requirements was $796,000.

                             During one period in 1986, the   Aviation Supply Office arbitrarily added
                             9 months to the administrative    leadtime standard for all purchases
                             because funds were available.    This action resulted in a 25-percent
                             increase in purchases reviewed    by GAO.




                             Page3                                    GAO/NSIAIMO-'78NavykadtimeForecast.s
                                                                                                           -
                          Executive   Summary




Production Leadtir ne     Unlike administrative leadtimes, production leadtimes are determined
                          on an item-by-item basis. However, the production leadtimes used in
Forecasts Are Not         requirement computations often did not realistically reflect the time
Adequate                  actually required to receive materials from contractors.

                          In determining production leadtime requirements, the Aviation Supply
                          Office automatically rejects actual leadtimes that are not within 80 to
                          125 percent of the leadtime on file. These parameters are intended to
                          ensure that atypical data do not influence the requirements forecast.
                          The Aviation Supply Office, however, does not review the leadtimes
                          that are outside of the parameters to determine if they actually are
                          representative.

                          In addition to historical information on experienced leadtimes, contrac-
                          tor estimates are to be used in forecasting production leadtime require-
                          ments according to DOD policy. However, officials at the Aviation Supply
                          Office stated that they receive contractor estimates on only about 5 per-
                          cent of the items they manage, and those that are obtained are not com-
                          pared with actual performance.

                          GAO’S  review of 21 randomly selected items showed that production
                          leadtime requirements for 17 items were overstated by $1.6 million and
                          requirements for 4 items were understated by $43,000. In one case, the
                          three most recently experienced leadtimes were below the 80 percent
                          parameter, which may indicate that the leadtime used in the require-
                          ments computation was too long.


Records Are Not Correct   The records used in the requirements determination process contained
                          numerous inaccuracies. For example, GAO’S analysis of 150 randomly
                          selected items showed that shipment and receipt records in the contract
                          status file did not agree for 122 (about 80 percent) of the items. GAO
                          projects that about 4,200 items, involving purchases of $487.5 million,
                          had file discrepancies between shipments and receipts.

                          Incomplete and inaccurate inventory records further hamper leadtime
                          forecasting. W ithout reliable data, the Aviation Supply Office does not
                          have reasonable assurances that the procurement system is adequately
                          protected from waste, fraud, and abuse.




                          Page 4                                  GAO/NSIAD-90-78   Navy Leadtime   Forecasts
                       Executive   Summary




                   A
                       GAO   recommends that the Navy implement procedures to more accu-
Recormnendations       rately determine both administrative and production leadtime require-
                       ments. Specifically, GAO recommends that administrative leadtime
                       requirements be based on actual experience for individual items and
                       that additives to requirements be fully justified. Further, GAO recom-
                       mends that production leadtimes that are outside of established parame-
                       ters be reviewed, contractor estimates be obtained and compared with
                       actual performance, and complete and accurate records be maintained
                       on administrative and production leadtimes.


Agency Comments        I) and acknowledged that improvements were needed in forecasting
                       leadtimes and maintaining accurate records, MODnoted ongoing or
                       planned corrective actions to more accurately determine leadtime
                       requirements and improve data accuracy. Data accuracy problems will
                       be targeted as an issue for review in fiscal year 1991 under the Federal
                       Manager’s Financial Integrity Act.

                       DOD   did not agree with GAO'S emphasis on contractor estimates in fore-
                        casting production leadtimes. GAO recognizes that these estimates are
                       just one of several sources of data but continues to believe that they
                        should be routinely obtained to help ensure that the most accurate pro-
                        duction leadtime forecasts are used in determining requirements.




                       Page6
Contents


Executive      Summary                                                                                    2

Chapter 1                                                                                                 8
Introduction             Objectives, Scope, and Methodology                                               9

Chapter 2
Administrative           ALT Requirements Have Fluctuated Considerably                                   11
                         ALTs for Individual Items Deviate Substantially From                            12
Leadtime Estimates           Standards
Can Be Improved          ALT Additive Increased Purchases                                                14
                         Conclusions                                                                     15
                         Recommendations                                                                 15
                         Agency Comments and Our Evaluation                                              15
                                                                                                     -
Chapter 3                                                                                                17
Production Leadtime      PLT Requirements Are Not Accurate                                               17
                         Leadtimes That Exceed Parameters Are Not Reviewed                               18
Estimates Can Be         Greater Use Can Be Made of Contractor Estimates                                 19
Improved                 Historical Records Are Not Complete                                             20
                         Conclusions                                                                     22
                         Recommendations                                                                 23
                         Agency Comments and Our Evaluation                                              23

Appendixes               Appendix   I: Comments From the Department of Defense                           26
                         Appendix   II: Major Contributors to This Report                                44

Tables                   Table 1.1: Procurement Leadtime Requirements for                                  9
                              Consumable Items
                         Table 2.1: Chronology of Administrative   Leadtime                              11
                              Standards
                         Table 3.1: GAO Projection of Discrepancies in the                               21
                              Contract Status File




                         Page 6                                GAO/NSL4D-90-78 Navy Leadtiie   Forecasts
Contents




Abbreviations

ALT        administrative leadtime
As0        Aviation Supply Office
DOD        Department of Defense
GAO        General Accounting Office
PLT        production leadtime

                                                                                     i
Page 7                                 GAO/‘NSIAMJO-78   Navy Leadtime   Forecasta   I
                                                                                     1
Chapter 1

Introduction


               The Department of Defense (DOD) estimates that parts, such as brakes,
               pumps, and compressors, costing $40 million are needed each day to sat-
               isfy military customers while awaiting additional deliveries for its
               inventories. Reducing the time to order and receive such spare parts can
               reduce the amount of material needed in inventory. This reduction can
               lessen the risk of unneeded inventories and can promote increased
               responsiveness to the operating military forces.

               The Naval Supply Systems Command administers the Navy supply sys-
               tem and provides supply management policies and procedures to its
               inventory control points. The Aviation Supply Office (ASO) is the Navy’s
               inventory control point for aviation material. As such, it is responsible
               for determining how many aviation spare parts are needed. ASO has
               developed standardized methods for determining how much stock is
               needed for initial provisioning, replenishment or peacetime operations,
               and mobilization for war.

               When determining requirements for replenishment inventories, the
               Navy considers economic order quantities and procurement leadtime
               and safety levels. The procurement leadtime level is the amount of
               material needed to meet normal demand during the time required to
               order and receive additional inventory. A safety level is the amount
               needed to meet fluctuations in demand and leadtime.

               The procurement leadtime level is further divided into administrative
               leadtime (ALT) and production leadtime (PLT). AI,T is the amount of mate-
               rial needed to meet demands from the time a need is determined until a
               contract is awarded to purchase the material+ PLT is the amount of mate-
               rial needed to meet demands during the time from the award of a con-
               tract until receipt of the material.

               The Navy maintains wholesale inventories at various stock points to fill
               requisitions from customers worldwide. These inventories include two
               types of material: consumables and repairables. Consumables are indi-
               vidual parts or assemblies that are disposed of when replaced.
               Repairables are components or assemblies that are returned to the sup-
               ply system to be repaired when replaced. For fiscal years 1986 to 1990,
               annual procurement leadtime requirements for aviation consumable
               items ranged from $2.1 to $2.9 billion, as shown in table 1.1.




               Page 8                                  GAO/NSIAD-90.78NavyLeadtimeForecasts
                                    Chapter 1
                                    Intraduction




Table 1 .l: Procurement Leadtime
Requirements for Consumable Items   Dollars In milllons                    I_-.-.____~
                                    Year                                                   ALT               PLT           Total
                                    1986                                                 $482.2      7gjiGY---         $2,122.6
                                                         -_     -   ~-~-
                                    1987                                                  570.4          1,736.6        2,307.O
                                                      .--~-
                                    1988                                                  882.3         1,651.3         2,533.6
                                                                  ~---                      -___,-
                                    1989                                                 1,288.8        l&24.3          2,913.i
                                                          .--   ~-.-                         ~___-
                                    1990                                                 1,160.O         1,552.a--      2,712.8


                                    Inaccurate leadtimes can have significant adverse effects, Overstated
                                    leadtimes can cause added investment for larger stock levels, greater
                                    chances of buying excess material, and increased termination costs if
                                    requirements change. Understated leadtimes can cause material
                                    shortages and reduced readiness of the units needing the material.


                                    Our objectives were to evaluate ASO'S requirements determination pro-
Ubjectives, Scope, and              cess for procurement leadtimes and to identify aspects that could
Methodology                         improve their accuracy.

                                    Between May 1989 and November 1989, we held discussions and col-
                                    lected information at the Office of the Secretary of Defense, the Naval
                                    Supply Systems Command, and ASO. We reviewed DOD guidance on defin-
                                    ing and developing procurement leadtime requirements and the Navy’s
                                    implementing policies, procedures, and practices. We selected Navy avi-
                                    ation consumable items for review because of the value of their leadtime
                                    requirements.

                                     To evaluate the reasonableness of results of ASO’s procurement leadtime
                                     process and related policies, we sampled and analyzed requirement eom-
                                     putations and available asset records ASO maintained. We compared
                                     leadtime estimates used by MO in requirement computations to lead-
                                     times actually experienced. We then recomputed the requirements using
                                     the actual data and compared the results to the requirements computed
                                     by using ASO'S estimates. We randomly sampled 150 consumable items
                                     from a universe of 6,109 consumable items, with purchases valued at
                                     $788 million, where ASO'S automated files indicated that supply con-
                                     tracts had been awarded and that some, but not all, of the material had
                                     been either received by the ru’avy or shipped to the Navy by a contrac-
                                     tor. We used this factor because ASO considers that leadtime ends when
                                     each consignee under a contract receives the first shipment. Primarily




                                     Page 9                                     GAO/NSIAD-90-78        Navy Leadtime   Forecasts
Chapter 1
Mroduction




due to inaccuracies in ASO’S contract status file, we obtained needed
requirement data on only 2 1 items.

For an additional   5,893 items, we used a computer analysis to compare
ALT estimates used by ASO in requirement computations to actual experi-
enced ALTS. The 5,893 items represent virtually all items where suffi-
cient data were available in ASO’S active files to compute the actual ALTS.

We used the same computer programs, reports, records, and statistics
the Navy used to manage inventories, make decisions, and determine
requirements. We did not independently determine the reliability of all
of these sources. However, we did assess the accuracy of ASO'S contract
status file by comparing data contained in the contract status file to
data maintained in other ASI files. We estimated the value of discrepan-
cies in the contract status file. We computed these estimates at a 95-
percent level of statistical confidence.

We performed    our review in accordance with generally accepted govern-
ment auditing   standards.




Page 10                                 GAO/NSZAD-90-78   Navy Leadtime   Fonxasts
-Chapter 2*

A dministrative Leadtirne Estimaks Can
l3e Improved

                                            Although ND policy requires that ALT be based on historical experience,
                                            A,SOgenerally has not used actual experience in determining ALT require-
                                            ments for individual consumable items. Instead, ASO has grouped items
                                            into a few categories and used the same standard ALT for all individual
                                            items within a group.

                                            These standards may reflect the average ALT experienced for an entire
                                            group, but they do not reflect the differences in individual leadtimes
                                            within a group. Our analysis of 2,467 items purchased during a recent
                                            period showed that actual ALT for 35 percent of the items deviated from
                                            the standards by more than 6 months.

                                            Our review of 21 randomly selected items showed that ALT requirements
                                            for 15 items were overstated by $645,000 and ALT requirements for 6
                                            items were understated by $796,000. During one period, ASO exacer-
                                            bated the situation by adding 9 months to the standard because funds
                                            were available.


                                            DOD policy, as promulgated    in DODInstruction 4140.55, requires that ALT
ALT Requirements                            be based on historical information that has been collected for represen-
Have Fluctuated                             tative procurements. However, ASO has not based ALT requirements on
Considerably                                actual experience for individual items. Instead, ASO has grouped the
                                            approximately    162,000 individual items into a few broad categories
                                            (such as all competitive procurements or noncompetitive procurements
                                            valued at between $25,000 and $99,999) and used the same standard
                                            ALT for all items within a group. These standards have fluctuated con-
                                            siderably over the past several years, as shown in table 2.1.
Table 2.1: Chronology   of Administrative
Leadtime Standards                          Date                  ALT standard                       ~      _.-..     ~____
                                             12/l 0185     __..   Minimum of 273 days for all items
                                                                                                         ~____~ --
                                            j/iap3                MinImum of 394 days for all items
                                            7/25/86               One-time 273 day additive to 394 day minlmum, or 667 days for
                                            _.___-                consumable    items only.
                                            12/i i/87             MInImums of 333,364, and 394 days, depending         upon the type of
                                                          -.~     material and extent of competition.
                                            8/04 /I38             Maximums of 182,273,333,3&I,        and 394 days, depending    upon
                                                                  the
                                                                  .-. type of material and extent of competitlon.
                                            12/20/88              Maximums of 165 days for all competitive procurements       and 128,
                                                                  143, 228, and 291 days for noncompetitive     procurements,
                                                                  dependkq    upon their values.




                                            Page 11                                       GAO/NSIAD-90-78     Navy Leadtime   Forecasts
                          Chapter 2
                          Administrative   Leadtime   Estimates   Can
                          Be Improved




                          Documents were not available at the time of our review to give the ratio-
                          nale for the earlier standards, ASO officials told us that they used stud-
                          ies, estimates, and experience in developing the standards. For example,
                          they stated that the increases in ALT standards for 1985 and 1986 were
                          caused by increases in the number of contracts awarded competitively
                          and decreases in the number of unpriced basic ordering agreements.
                          Instead of unpriced orders, contracts were not awarded until they had
                          been priced out-this    increased leadtime requirements.

                          Several reasons for subsequent decreases in ALT standards were cited in
                          various documents and discussions. A memorandum setting the Decem-
                          ber 1987 standards stated they were based on experience gained in deal-
                          ing with increased competition and firm-fixed price contracts. The
                          August 1988 standards were based on a study of actual ALTS experi-
                          enced for groups of items. Other reasons given by MO officials for the
                          decreases in ALT standards include

                      l establishing a monitoring system to track the performance of inventory
                        managers and contracting personnel;
                      . performing some functions concurrently instead of sequentially;
                      l using more flexible contractual processes, such as including option
                        clauses to purchase larger quantities; and
                      + adjusting the workloads of contracting personnel so that they are only
                        given the number of purchase requests that can be realistically handled.

                          The officials stated that the December 1988 changes were based prima-
                          rily on the ASO commanding officer’s prior experience working with the
                          Defense Logistics Agency.


                          Our analysis of 5,893 purchases, valued at $119.4 million, showed that
ALTs for Individual       average ALTS at tio have declined since 1985. However, further analysis
Items Deviate             showed that ALTS for individual items have varied substantially from
Substantially From        these averages, as well as from the ALT standards used in the require-
                          ment computations.
Standards
                          Of the 5,893 purchases, 2,467 purchases were initiated between July 18,
                          1986, and December 10, 1987. They illustrate the extent of the variance.
                          During this period, the average ALT was 418 days and the standard was
                          a minimum of 394 days for all items. We found that the actual ALT for
                          863 purchases varied from the standard by at least 180 days. In other
                          words, MO either overstated or understated ALT requirements by 180
                          days or more in 35 percent of the purchases we analyzed.


                          Page 12                                       GAO/NSIAD90-78   Navy Leadtime   Forecasts
Chapter 2
Administrative   Leadtime   Estimates Can
Be Improved




To further quantify the extent of individual deviations from the
standards, we reviewed data on 2 1 randomly selected consumable items.
We found that the average ALT exceeded the standards by as much as
601 days for 15 items and that 10 of them exceeded 180 days. The aver-
age ALT was less than the standard by as much as 749 days for the other
6 items. We substituted the actual ALTS for the standards used in the
requirement computations and determined that requirements for 15
items were overstated by $645,000 and requirements for 6 items were
understated by $796,000. The following examples illustrate the poten-
tial impact of not using representative ALTS in requirement
determinations.

ASO used an ALT of 394 days in computing   a requirement for 160 canopy
glass assemblies (NSN-1560-00-402-8654)     for the A-6 aircraft. ASO initi-
ated a purchase action for the 160 assemblies on April 2, 1987, and
awarded a contract on October 2, 1987. Therefore, the actual ALT was
183 days. The difference of 211 days in ALT equates to the purchase of
24 additional assemblies with an estimated contract value of $121,920.
On the basis of an ALT of 183 days, the requirement would have been
reduced from 160 to 136 assemblies.

In another case, AS0 used an ALT of 684 days in computing a requirement
for 523 digital microcircuits (NSN-5962-00-225-0472)    for the F-14 air-
craft. ASO initiated a purchase action for the 523 microcircuits on July
29, 1985, and awarded a cont,ract on April 2, 1986. Therefore, the actual
ALT period was 247 days. The difference of 437 days in ALT equates to
the purchase of 192 additional microcircuits with an estimated contract
value of $96,960. On the basis of an ALT of 247 days, the requirement
would have been 33 1 microcircuits.

As a final example, ASO used an ALT of 751 days in computing a require-
ment for 60 breather adapters (NSN-1560-00-970-9693)       for the A-6 air-
craft AS0 initiated a purchase action for the 60 adapters on September
15, 1986, and awarded a $3,960 contract on February 12, 1987. There-
fore, the actual AI=I' was 150 days. No purchase of breather adapters
would have been necessary at that time because with an ALT of 150 days
sufficient assets would have been available to meet the requirements.

Supply officials stated they grouped items because ALT requirement
determinations for individual items would have required input from
inventory managers on each item and they did not want to do that. How-
ever, our discussions with DOD and Navy officials indicate that it is feasi-
ble to make ALT forecasts on an individual item basis. According to these


Page 13                                     GAO/NSIALMO-78   Navy   Leadtime   Forecasts
                      Chapter 2
                      Administrative   Leadtime   Estimates Can
                      Be Improved




                      officials, inventory managers in both the Army and the Air Force review
                      historical data and make ALT forecasts for individual items. They also
                      stated that computer equipment was in place for individual ALT determi-
                      nations in the Navy. We noted that ASO'S automated files contained the
                      data entries needed to compute ALTS for individual items.


                      On July 25, 1986, ASO added 273 days to the existing standard of 394
ALT Additive          days for consumable items. ASO did this in order to purchase additional
Increased Purchases   stock. Specifically, ASO had fiscal year 1986 funds available that enabled
                      it to satisfy requirements for future periods. By using the additive, ASO
                      raised the reorder levels, thereby generating purchases that otherwise
                      would not have been made or would have been made for lesser quanti-
                      ties. The additive was a one-time occurrence and was eliminated after
                      the purchases it generated were initiated.

                      Since requirements data on 64 consumable items affected by this addi-
                      tive were available from our prior review of operating levels, we used
                      the data to determine the effect of the additive on ASO'S purchases. We
                      compared the actual purchase values for the 64 items with what the
                      values would have been without the additive. The comparison showed
                      that the total purchase value would have been reduced by $2.1 million
                      (from $10.6 million to $8.5 million). By eliminating the additive, 50
                      purchases would have been for lesser quantities and 14 purchases
                      would not have been made. The following examples illustrate the impact
                      of the additive.

                      ASO used an ALT of 667 days in computing   a requirement for 349 direct
                      current motors (NSN-6105-00-858-6873)    for the H-3 helicopter. Since
                      138 motors already were available, ASO initiated a purchase for 2 11
                      additional motors. If 273 days had not been arbitrarily added to the 394-
                      day ALT standard, ASO would have purchased 61 fewer motors and saved
                      $52,338.

                      In another case, ASO used an ALT of 667 days in computing a requirement
                      for 240 toggle switches (NSN-5930-01-032-0644)    for the H-l helicopter.
                      Since 161 switches already were available, ASO initiated a purchase for
                      79 additional switches. If 273 days had not been arbitrarily added to the
                      394-day ALT standard, AS0 would not have had to purchase any switches
                      and it would have saved $19,355.




                      Page 14                                     GAO/NSIAIHO-78   Navy Leadtime   Forecasts
                      Chapter 2
                      Administrative   Leadtime   Estimates Can
                      Eie Improved




                      ASO generally has not used actual experience in determining ALT require-
Conclusions           ments for individual consumable items, even though DODpolicy requires
                      that ALT be based on historical information for representative procure-
                      ments. Instead, ASO has grouped items into a few categories and used the
                      same standard ALT for all items within a group. These standards may
                      reflect the average ALT experienced for an entire group, but they do not
                      reflect the divergence in individual leadtimes within the group.

                      We believe that ALTS used in requirement determinations should be
                      based on actual experience for individual items. The Army and Air
                      Force do this and it is feasible for the Navy to do it. To the extent that
                      representative, past experience is used to develop current requirements,
                      those requirements will provide a more accurate and realistic basis for
                      procurement decisions.

                      We also believe that the use of additives to artificially increase ALT
                      requirements should be limited, reasonable, and fully justified. The
                      availability of extra funds does not seem to us to be sufficient reason to
                      increase requirements in order to prematurely purchase additional
                      stocks.


                      We recommend that the Secretary of the Navy direct the Commander,
Recommendations       Naval Supply Systems Command, to implement procedures that more
                      accurately determine administrative leadtime requirements. Specifi-
                      cally, we recommend that the Commander

                      base administrative leadtime requirements on representative actual
                      experience for individual items rather than on standard leadtimes for
                      groups of items,
                      ensure that additives to administrative leadtime requirements are rea-
                      sonable and fully justified, and
                      review historical data on administrative leadtime requirements for com-
                      pleteness and accuracy.


                      DOD generally  agreed with our recommendations and stated that ASO had
Agency Comments and   implemented procedures to more accurately determine ALT require-
Our Evaluation        ments. One procedure changes the data base to liberally accept shorter
                      ALTS while limiting the acceptance of longer ALTS. Another procedure
                      improves the timeliness of ALT forecast updates by recording actual ALTS
                      shortly after contract award rather than waiting until contract deliv-
                      eries occur.


                      Page 15                                     GAO/NSlADSO-78   Navy Leadtime   Forecasts
Chapter 2
Administrative   Leadtime   Estimates Can
Be Improved




DOD did not agree that the July 1986 additive to ACT requirements was
based on available funding. DOD stated that the additive was made to
preclude a potential stockout position during the transition from an
unpriced purchase order environment         to a more time consuming priced
order environment.

We continue to believe that the additive was principally driven by avail-
able funding as evidenced by the fact that AS0 eliminated the additive
immediately after the purchases generated were initiated. Also,
although ASO’S initial budget execution plan for fiscal year 1986 called
for obligating 20 percent of the budget in the fourth quarter, ASO actu-
ally obligated 40 percent during the fourth quarter. Furthermore, the
additional time required to price out orders already had been accommo-
dated in earlier standards. On December 18, 1985, ASO directed an ALT
minimum of 273 days for all items to cover the proposed reductions in
unpriced orders. On July 18, 1986, AS0 increased the minimum to 394
days for essentially the same reason. The latter date was just one week
before ASO provided for the additive.




Page 16                                     GAO/NSLAD-90-78   Navy Leadtime   Forecasts
Production Leadtime Estimates Can
Be Improved

                         According to DOD policy, ASO is to use historical information and cOnh2C-
                         tor estimates in establishing realistic PLT requirements. Our review indi-
                         cated that the PLTS used often did not reflect the time actually required
                         to receive materials from contractors. This was because ASO did not (1)
                         review historical leadtimes that were outside of established parameters,
                         (2) routinely obtain contractor estimates or compare those that had been
                         obtained to actual contractor performance, and (3) maintain complete
                         and accurate historical file data.

                         As with ALT, inaccurate PLTS can have significant adverse effects. Over-
                         stated PLT requirements can generate excess stocks and understated PLT
                         requirements can cause stock shortages. Our review of 21 randomly
                         selected consumable items showed that PLT requirements for 17 items
                         were overstated by $1.6 million and that PLT requirements for 4 items
                         were understated by $43,000.


                         To make realistic forecasts and ensure that neither too many nor too
PLT Requirements   Are   few spare parts are purchased, DOD policy states that PLTS used in the
Not Accurate             requirements determination      process should be based on estimates from
                         contractors or historical information that has been collected for repre-
                         sentative procurements. In line with this policy, ASO collects data on
                         actual experienced PLTS and contractor estimates. However, we found
                         that the PLTS used in the requirements determination     process often did
                         not realistically reflect the time actually required to receive materials
                         from contractors.

                         We reviewed the PLT requirements for 21 randomly selected consumable
                         items and found that the PLTS used in the requirements computation dif-
                         fered from the actual PLTS in every case. The PLTS for 17 items were
                         overstated by up to 472 days and the PLTS for 4 items were understated
                         by up to 631 days. We substituted the actual PLTS for those used in the
                         requirement computations and determined that, on total purchases of
                         $7.5 million, requirements for the 17 items were overstated by $1.6 mil-
                         lion and requirements for the 4 items were understated by $43,000. The
                         following examples illustrate the potential impact of inaccurate 1'1,~s.

                         On November 24,1986, AS0 used a PLT of 573 days in computing a
                         requirement for 10,488 compressor blades (NSN-2840-00-810-9309)       for
                         the J-52 engine, even though deliveries under the two most recent con-
                         tracts were 200 and 349 days. On September 25, 1987, ASO exercised a
                         contract option clause for an additional 2,622 blades and awarded a con-
                         tract for 13,110 blades. On February 19, 1988, the .Jacksonville Naval


                         Page17                                   GAO/NSIAD%-78Navy   LeadtimeForecasts
                    Chapter 3
                    Pruduction kadtime    Estimates   Can
                    Be Improved




                    Supply Center received the first shipment; therefore, the PLT for the
                    blades was 147 days. The difference of 426 days between the estimated
                    PLT of 573 days used by ASO and the actual FLT of 147 days equates to
                    the purchase of 2,659 more blades (at an estimated contract value of
                    $446,712) than was required.

                    Using a PLT of 573 days in the requirements computation contributed to
                    a subsequent excess of blades. In May 1988, ASO determined that 4,628
                    excess blades were under contract and notified the contractor to cease
                    work because the contract was being terminated. In June 1989, the con-
                    tract amount was decreased by $779,448 to reflect the reduced
                    quantities.

                    In another case, on January 15,1986, ASO used a PLT of 482 days in com-
                    puting a requirement for 356 shroud assemblies (NSN-1560-00-064-
                    9374) for the A-6 aircraft. No information on prior contracts was in
                    ASO’S contract status file. On January 27, 1987, ASO awarded a contract
                    for 356 assemblies and on February 10,1988, the Oakland Naval Supply
                    Center, the second of two consignees under the contract, received 12 of
                    these assemblies. Therefore, the PLT for the assemblies was 379 days.
                    The difference of 103 days between the estimated PLT of 482 days used
                    by ASO and the actual PLT of 379 days equates to the purchase of 39
                    more assemblies (at an estimated contract value of $14,664) than was
                    required.

                    Th? following        sections discuss some of the reasons for inaccurate PLT
                    requirements.


                    When forecasting future PLT requirements, ASO officials told us that they
Leadtimes That      did not review actual experienced PLTS or contractor estimates that
Exceed Parameters   exceeded established parameters. In determining PLT requirements, ASO
Are Not Reviewed    uses a process that automatically rejects variations from the PLT forecast
                    if the variations are outside of prescribed parameters. The upper param-
                    eter is 125 percent of the PLT currently in file and the lower parameter is
                    80 percent of the PLT currently in file.

                    This forecasting method is intended to ensure that atypical data from
                    actual experience or contractor estimates do not influence the require-
                    ments forecast. Data that are within the parameters are weighted to
                    place major emphasis on more recent data and lesser emphasis on older
                    data.



                    Page 18                                      GAO/NSIAD9O=78   Navy Lemltime   Forecasts
                     Chapter3
                     ProductionLeadtimeEstimatesCan
                     Be improved




                     Although ASO’S automatic process has the advantage of isolating appar-
                     ent atypical PLTS, ASO does not review the PLTs that are outside of the
                     parameters to determine if they are truly nonrepresentative.    Experi-
                     enced PLTS that fall within the 80 and 125 percent parameters are
                     accepted. Data on the other PLTS are collected separately in the ASO corn--³
                     puter system but are not reviewed.

                     Of the 21 consumable items we reviewed, 14 items had experienced PLTS
                     that were outside of the parameters. For example, in the case of the
                     previously discussed compressor blades, ASO’S files continue to show a
                     PLT of 573 days because the latest experienced PLT of 147 days and the
                     two prior PLTSof 200 and 349 days were below the 80 percent parameter
                     and therefore were not reviewed. Similarly, ASO’S files continue to show
                     a PLT of 482 days for the shroud assemblies because the latest experi-
                     enced PLT of 379 days was below the 80 percent parameter and, there-
                     fore, was not reviewed.

                     DOD  policy states that when experienced PLTS or contractor estimates are
                     determined to be nonrepresentative    of future performance, they should
                     be excluded from normal PLT development. The policy also states, how-
                     ever, that the data should be retained in the procurement history files.
                     Unless PLTS that are outside of the parameters are reviewed, AS0 will not
                     know if they are representative or nonrepresentative.


                     ASO generally does not obtain contractor estimates before determining
Greater Use Can Be   PLT requirements.  When estimates are received, they are not compared
Made of Contractor   with actual performance. DOD policy states that such estimates and com-
Estimates            parisons shall be used so that the most accurate PLT forecasts can be
                     made. AS0 obtains and uses PLT estimates from only five major contrac-
                     tors. ASO does not maintain statistics on the number of individual con-
                     tract estimates received, but ASO officials estimate that the estimates
                     apply to only about 5 percent of the items in inventory.

                     If contractor estimates were routinely obtained and used in calculating
                     PLT requirements, more accurate PLT forecasts could result. For example,
                     on May 3, 1989, AS0 used a PLT of 473 days in computing a requirement
                     for 29 engine shields (NSN-2840-01-162-1143)    for the F-402 engine. We
                     asked a contractor representative for a PLT estimate who stated that,
                     under normal conditions, first delivery would take place within 57
                     weeks, or 399 days after ASO'S order was received. The difference of 74
                     days between the estimated PLT used and the contractor’s estimate
                     equates to the purchase of four extra shields, with an estimated value of


                     Page19                                   GAO/NSIAD90-78NavyLmdtimeForecasts
                         Chapbr3
                         production   Ixiadtitne   Ehdimam   Can
                         Be Improved




                         $7,960, If actual performance was subsequently compared with the con-
                         tractor estimate and the PLT used in the requirements computation,
                         future PLT requirements would be more accurate. Unfortunately,     Aso
                         does not have a procedure for making these comparisons.

                         One area where positive action has been directed is in the area of con
                         tract negotiations. DOD has been concerned about the length of leadtimes,
                         and in May 1989, held a major conference on reducing leadtimes. Nearly
                         100 top military and civilian officials from the procurement commands
                         of the military services and the Defense Logistics Agency attended. One
                         of the management concepts identified for possible use non-wide was to
                         request shorter PLTS in contractor solicitations and press for continued
                         reductions. In the past, Navy procurement policy was not clearly ori-
                         ented toward negotiating reductions in PLT. In May 1989, the Secretary
                         of the Navy reported to DOD that this policy had changed, and the Navy
                         now is negotiating PLT reductions with contractors.


                         We found numerous inaccuracies in the ASO records used in the PLT
Historical Records Are   requirements determination process. Incomplete and inaccurate records
Not Complete             make realistic PLT forecasts a very difficult task. W ithout reliable data,
                         ASO does not have reasonable assurances that the procurement system is
                         adequately protected from waste, fraud, and abuse.

                         ASO’S automated  contract status file is a permanent  record of shipments
                         by contractors and receipts by Navy activities. ASO uses this file to
                         research contract shipment and receipt data. We examined the April
                         1989 contract status file to determine the number of consumable items
                         with partial shipments or receipts. The file indicated that 6,109 items,
                         valued at $788 million, fell into these categories.

                         We randomly sampled 150 of these items for review and asked ASO to
                         provide requirements determination documents. ASCJcould provide docu-
                         ments on only 2 1 items. Requirements documents on the other 129 items
                         were not available, primarily because ASO officials indicated that all
                         material had been delivered and the contracts were completed.

                         Because the contract status file indicated that none of the contracts for
                         the 150 items were complete, we analyzed the file data in more detail.
                         Our analysis showed that shipment and receipt records did not agree for
                         122 (about 80 percent) of the 150 items. On the basis of our analysis, we
                         estimate that 4,154 items, involving purchases of $487-5 million, had file
                         discrepancies between shipments and receipts.


                         Page20                                    GAO/NSIAD9W78NavyLeadtimeForecasts
                                            Chapter 3
                                            Production Leadtime       Estimates   Can
                                            Be Improved




                                            For example, in October 1987, ASO awarded a contract for 85 electromag-
                                            netic relays (NSN-5945-OO-450-4679),     valued at $10,156, for the H-l
                                            helicopter. The contract status file indicated that the contractor had
                                            shipped all 85 relays in February 1988, but the file also indicated that
                                            none had been received as of April 1989. The inventory manager for the
                                            relays told us that the contract was completed in February 1988 and
                                            that requirements documents were no longer available.

                                            Our projection of the types of discrepancies in the contract status file
                                            based on our sample of 150 out of a universe of 6,109 items is summa-
                                            rized in table 3.1. Our sample provides a 95-percent level of statistical
                                            confidence.

Table 3.1: GAO Projection of
Discrepancies in the Contract Status File   Dollars in milllons
                                                                                                                     Number of           Value of
                                            Type of discrepancy           .---                                           items        purchases
                                            Quantities were       shipped but none were received               ~_-        1,507             $25.1
                                            Quantities were       received but none were shipped                          1,670             2252
                                            Quantities were       shipped and quantities were received   but
                                              totals did not      agree
                                                                  ___.~-.~-                                                977              237.2
                                            Total discrepancies                                                          4,154            $487.5


                                            In addition to the contract status file, ASO maintains a due in file. The
                                            latter file is used to automatically compute the number of days in PLT.
                                            Entries in the due in file are temporary, and the file is purged when
                                            receipts equal the total quantities ordered under a contract. Permanent
                                            receipt data are retained in the contract status file because this file is
                                            supposed to be updated with receipt data at the same time the receipts
                                            are recorded in the due in file.

                                            To determine if the contract status file and the due in file were in agree-
                                            ment, we compared the content of the files for the 150 sample items. Our
                                            comparison identified differences on 74 items. For these items, the con-
                                            tract status file showed outstanding receivables, but the due in file
                                            showed no receivables. ASO officials told us that inventory managers
                                            probably had deleted the receivables in the due in file but had not
                                            updated the contract status file.

                                            ASO officials stated that a major reason why accurate records are not
                                            maintained is that ASO’S data files are fed information from a variety of
                                            sources, and this leads to errors, mismatches, nonpostings, and other



                                            Page 21                                                GAO/NSlALMO-78     Navy Leadtime    Forecasts
              Chapter 3
              Production    kadtime   Estimates   Can
              Be Improved




              problems, Nevertheless, the discrepancies within and between ASO'S files
              indicate to us that internal controls over file data need to be improved.

              Internal controls are essential elements of effective inventory manage-
              ment. When properly implemented, they provide reasonable assurance
              that (1) reliable data are obtained, maintained, and recorded, (2) assets
              are protected from waste, fraud, and abuse, and (3) resources are used
              in accordance with applicable laws, regulations, and policies. The data
              accuracy problems we found indicated that the Navy did not have ade-
              quate internal controls over this important data.

              The Federal Manager’s Financial Integrity Act of 1982 (3 1 U.S.C.
              3512(b)) requires that agency internal control systems be periodically
              evaluated and that agency heads provide annual reports to the Presi-
              dent and the Congress that state whether these systems comply with the
              objectives of internal controls set forth in the act and with the standards
              prescribed by the Comptroller General, When systems do not comply,
              agency reports must identify the weaknesses involved and describe the
              plans for corrective action. We reviewed these reports to determine if
              the Navy had identified significant weaknesses in internal controls over
              kso’s file data discussed herein and found that the Navy had not done
              so.


              ASO does not review HIT data that the computer system automatically
Conclusions   screens out if they are outside of established parameters. kso needs to
              review the screened out data to determine if the PLTS are truly nonrepre-
              sentative or if conditions have changed so that they now are
              representative.

              ASO receives PLT estimates from only a few contractors and those that
              are received are not compared with actual performance. Obtaining con-
              tractor estimates and comparing them with actual performance would
              be useful in updating PLT forecasts and would be in accordance with DOD
              policy.

              Our review indicated that completeness and accuracy of PLT data in the
              permanent files is a serious problem. Shipment and receipt information
              currently in the contract status file is highly inaccurate and does not
              provide an adequate data base for research when updating PLT fore-
              casts. W ithout reliable records and adequate internal controls, ASO does
              not have reasonable assurances that assets are protected from waste,
              fraud, and abuse.


              Page 22                                   GAO/NSIAD-90-78   Navy Leadtime   Forecasts
                          Chapter 3
                          Production Leadtime   Estimates   Can
                          Be Improved




                          Improvements in reviewing PLTs that are not within established parame-
                          ters, obtaining and comparing contractor estimates, and maintaining
                          complete and accurate historical files would help ensure that better
                          forecasting data are available for use in determining PLT requirements.
                          More accurate requirement determinations would reduce the chances of
                          generating excess stocks because of overstated PLT requirements or
                          incurring stock shortages because of understated PLT requirements.

                  A
                          We recommend that the Secretary of the Navy direct the Commander,
Recommendations           Naval Supply Systems Command, to implement procedures to ensure
                          that more accurate production leadtime forecasting data are available
                          and used in making requirement determinations. Specifically, we recom-
                          mend that the Commander

                      . review experienced leadtimes and contractor estimates that are outside
                        of established parameters to determine if they are representative of cur-
                        rent conditions;
                      . routinely obtain contractor estimates of production leadtime and com-
                        pare the estimates with actual contractor performance;
                      . improve the completeness and accuracy of production leadtime data in
                        the historical files and establish the internal controls needed to ensure
                        the accuracy of these files, including the reconciliation of shipment and
                        receipt data in the contract status file; and
                      . target data accuracy problems as an issue for review in the Federal
                        Manager’s Financial Integrity Act assessments.


                          DOD stated that as part of an overallDOD initiative to reduce procure-
Agency Comments and       ment leadtimes, increased emphasis has been placed on determining the
Our Evaluation            best available PLTS and accurately recording realistic PLTS. DOD acknowl-
                          edged that the degree of variation between actual PLTS and those used in
                          requirement determinations was a valid concern.

                          W ith regard to our data accuracy recommendations, DOD stated that
                          data processing modernization efforts under an ongoing major software
                          resystemization project will ensure that assets are properly reflected in
                          all accountable records, As an interim measure, ASO is pursuing short
                          term diagnostic methods to alleviate the effects of current weaknesses.
                          DOD also stated that data accuracy problems will be targeted as an issue
                          for review in fiscal year 1991 under the Financial Integrity Act
                          assessments.



                          Page 23                                 GAO/NSIAD-W-78   Navy Leadtime   Forecasts
Chapter 3
Production Leadtime   Estimates Can
Be tmproved




DOD did not agree with our recommendation      to routinely obtain contrac-
tor estimates of PLT and compare the estimates with actual contractor
performance. DOD stated that, although contractor estimates are valua-
ble, they are just one of several sources of data used in validating PLT
forecasts and they can be unreliable.

We did not intend to imply that contractor estimates should be the sole
source for PLT forecasts. We agree with DODthat these estimates are just
one of several sources of data but continue to believe that they should
be routinely obtained to assure that the most accurate PLT forecasts are
used in determining requirements. We also agree that contractor esti-
mates can be unreliable and that is why we are recommending that they
be compared with actual performance.




                                                                                      D




                                                                                      ,




Page 24                                  GAO/NSIAIHO-78   Navy Leadtime   Forecasts
                                                       i




Page 25   GAO/NSL4IMlO-78   Navy hadtime   Forecasts
Appendix      I

Comments From the Departxnent of Defense


Note: GAO comments
supplementing     those In the
report text appear at the
                                                                 ASSISTANT     SECRETARY     OF DEFENSE
end of this appendix.                                                   WASHINGTON.   0 c. 20301-8000
                                        I




                                                                                                   APR 1 1 1990



                                 Mr. Prank       C. Conahan
                                 Assistant       Comptroller       General
                                 National       Security       and International
                                   Affairs   Division
                                 U.S. General Accounting               Office
                                 Washington,    DC 20548

                                 Dear Mr. Conahan:

                                        This is the Department       of Defense (DOD) response to the General
                                 Accounting    Office    (GAO) draft    report,  "NAVY SUPPLY: Procurement
                                 Leadtime Forecasting       Needs Improvement,"     dated January 30, 1990 (GAO
                                 Code 394277/OSD      Case  8233).     The DOD generally       concurs with both the
                                 findings    and recommendations       of the draft    report.

                                        The DOD agrees that improvements           are needed in the forecasting         of
                                 lead times.        To that end, the Navy Aviation          Supply Office    has
                                 adjusted     its lead time filter       parameters     both to reflect    actual   lead
                                 times more accurately         and to implement      the DOD-wide initiative      to
                                 reduce procurement        lead times.      The DOD also agrees that improvements
                                 in file     data accuracy     are required.      The ongoing Navy Resystemization
                                 effort    includes    improvements    in file    data accuracy    as a major goal.

                                      While the DOD regards       information       from contractors          as valuable  in
                                 determining    realistic   production      lead time estimates,            the Department
                                 does not agree with the proposed degree of reliance                      on contractor
                                 estimates.     As part of the overall         DOD Initiative         to reduce
                                 procurement    lead times,    including      production      lead times,       increased
                                 emphasis has been placed both on determining                 the best available
                                 production   lead times and accurately           recording      realistic      production
                                 lead times.

                                      The findings             and recommendations       are addressed      in greater detail
                                 in the enclosure.                The DOD appreciates       the opportunity     to comment on
                                 the draft   report.

                                                                                             Sincerely,

                                  Enclosure




                                            Page 26                                              GAO/NSIAIMO-78   Navy Leadtime   Forecasts
                                  Appendix I
                                  Comments From the Department             of Defense




                                             GAO DRAFT REPORT - DATED JANUARY 30, 1990
                                                   (GAO CODE 394277) OSD CASE 8233

                          "NAVY SupPLY:         PROCUREMENTLEADTIME FOREXASTING NEEDS IMPROVEMENT"

                                                  DEPARTMENT OF DEFENSE COMMENTS



                                                               FINDINGS

                          FINDING A: Backqround:                     Naw SuppIv Lead Time Forecastinq.                          The GAO
                          reported      that the Naval Supply Systems Command administers                                   the Navy
                          supply system, while the Aviation                       Supply Office            is the Navy inventory
                          control     point       for aviation         material.         The GAO observed that, in
                          determining        inventory         requirements,         the Navy considers              estimated
                          procurement        lead time, safety levels                   and economic order quantities.
                          The GAO expLained             that procurement           lead     time     level     is the amount of
                          inventory        needed to meet normal demand during                         the time required            to
                          order an'd receEve additional.inventory.                            The GAO also noted that the
                          procurement         lead    time     level    is further         divided       into administrative
                          leadtime      and production             lead time,         (The GAO further             explained       that
                          the amount, of matel;ial               needed.to       meet demands from the time a need is
                          determined        until     a contract        is awarded is called                 administrative          lead
                          time, while the amount of material                       needed to meet demands during the
                          time from the-award               of a contract         until      receipt        of the material          is
                          called     production          lead time.)         The GAO found that,                for FY 1986
                          through      EY 1990, annual procurement                    lead time requirements                 for
                          aviation       consumables          ranged from $2.1 billion                 to $2.9 billion            (as
                           shown in table 3.1 in the draft report).                             The GAO concluded that,
                          because the Aviation                Supply Office manages approximately                        162,000
                          aviation       items; determining             valid     requirements            is a significant
                           challenge.         The GAO further           concluded         that the sheer magnitude also
                          makes inventory            management vulnerable               to inefficlency            and waste.
Now on pp.2,8,   and 9.    fpp& l-2, pp. lo-lZ/GAO                 Draft Report)

                          DOD Response:        Concur.     The challenge       inherent      in determining       valid
                          requirements      for an inventory      of 171,000 aviation             consumable items
                          and 96,000 aviation        repairable     items is, indeed,            considerable.       In
                          meeting the challenge,         the Aviation        Supply Office        has instituted
                          procedures     armed at minimizing        inefficiency        and waste and ensuring
                          that requirements,      determinations       reflect     the maximum degree of
                          accuracy.      As a routine      part of business,         several      recurring    actions
                          allow item managers the opportunity               to validate        the data that is
                          critical     tp the requlrementa       determination        process,       including




                                   Page 27                                                      GAO/NSIAlHO-78         Navy Leadtime        Forecasts
        Appendix I
        Comments From the Department          of Defense




administrative          and production          lead time forecasts.         Supply Demand
Review is a computer program,                   run approximately       biweekly,     which
determines        on an item-by-item            basis how much and when to buy.              The
Stratification          process is a semiannual             program used to categorize
stock according           to the purposes          for which it is held and to develop
an estimate         of future      spares budget requirements.              Internal    audits of
procurements          and Requirements          Review Boards also provide           a safeguard
system to ensure accurate                 calculation      of requirements.
Implementation           of statistical         process controls       by the Aviation       Supply
Office       (commencing in March 1990) will                 result  in recommendations
targeted       at the specific          processes      causing delays in contract           awards,
thereby       enabling      future    procurements       to be based on highly         accurate
estimates        of administrative           lead time.

It is important          to recognize       that the management of any large
organization         entails    risk.     The whole point of the Aviation              Supply
Office      review process is to focus attention                on the most significant
risks.       With finite      resources,       it is impossible        to eliminate/prevent
all vestiges         of "vulnerability         to inefficiency       and waste" in any
operation       this large.         Improvement     is always possible.           There are,
however, appropriate            finite    limits    on the efforts       made, because
otherwise        the attempts       to eliminate      all vulnerability        risk becoming
inefficlent         and wasteful       in and of themselves.

FINDING 0: Administrative              Lead Time Requirements             Have Fluctuated
Considerablv.         The GAO reported         that, although         DOD policy        requires
that administrative          lead time forecasts           be based on historical
information        (which has been collected            for representative
procurements),        the Aviation       Supply office        generally      has not used
actual     experience      in determining        lead time requirements             for
individual       consumable items.          The GAO found that,           instead,        the
Aviation      Supply Office       groups its 162,000 different               stock items Into
only a few groups and uses the same standard                       lead time for all
individual       items within       a group.       The GAO observed that,            while these
standards      may reflect      the average lead time experienced                   for the
entire     group, they do not reflect              the differences        ln individual          lead
times within        a group.      The GAO also observed that these standards
have fluctuated         considerably       over the past several            years.         (The GAO
provided      a chronology      of the changes in administrative                  lead time
standards      in table 2.1, as well as the various                   expl.anations         for these
changes.)        The GAO reported        that,     according     to supply officials,              they
grouped items because,            if they computed administrative                 lead time
requirements        on the basis of individual             items, it would require               input
from inventory        managers.       The GAO noted, however, that the Army and



                                                    2




        Page 28                                                   GAO/NSlAD90-78       Navy Leadtime      Forecasts
                                      Appendix I
                                      Comments From the Department        of Defense




                              the Air Force estimate              administrative     lead time for individual             items.
                              In addition,        the GAO observed that computer equipment                    and the
                              necessary       data were in place for individual               administrative         lead time
                              determinations         by the Navy.          The G.&o concluded,     therefore,        that it is
                              feasible      for the Navy to also compute administrative                     lead time
                              requirements        on the basis of individual            items.     The GAO further
                              concluded      that administrative            lead times should be based on actual
                              experience       for individual         items --not collectively        by groups.         The GAO
                              also concluded          that,     to the extent that representative,              past
                              experience       is used to develop current            requirements,        those
                               requirements       will provide        a more accurate      and realistic         basis for
                              decisions.         In summary, the GAO concluded             that reducing         the time to
                              order and receive             items can reduce the amount of material                 needed in
                               inventory--which,            in turn,    can lessen the risk of unneeded
                               inventories       and can promote increased           responsiveness         to the operating
Now   on pp. 3,11,   12,and   Military      Forces.          (p. 4, pp. 15-18, p. 23/GAO Draft Report)
15.
                              DOD Response:         Concur.       The Aviation         Supply Office        has taken a number
                              of actions      in the past three years to stabilize                      and reduce
                              administrative        lead    time and, as a direct             result     of those actions,
                              the overall       average aciministratlve            lead time has been reduced from
                              360 days to 172 days.             The Aviation         Supply Office         has established
                              administrative        lead time goals, the purpose and intent                        of which were
                              to focus attention           at all levels       of the Aviation           Supply Office on the
                              internal     processing       time required        to award contracts.               The
                              administrative        lead time goals represented                 a significant         reduction
                              from the actual          experienced      administrative          lead time.         The
                              administrative        lead time goals were overlaid                  in the Master Item File
                              for each item, unless the existing                   file value was lower than the
                              goal, in which case the lower administrative                         lead time remained
                              unchanged.        Requirements       determination          calculations        were based on the
                              administrative         lead time goal or the lower experienced                       administrative
                              lead time.        Consequently,        the overall         administrative         lead time was
                              reduced as a result            of overlaying       the administrative             lead time target
                              in the Master Item File.               In addition,         the Aviation        Supply Office
                              reset the lead time filters               to ensure consideration               of reduced
                              administrative         lead time initiatives             in all future        computations        of
                              administrative         lead time.

                              It is certainly      true that administrative     lead time requirements     have
                              fluctuated    considerably,     particularly  over the period     of 1984 to
                              1987, the interval       durinq which the procurement     actions    studied by
                              the GAO were init iated.       -Major changes in contracting      pal icies and




                                                                                3




                                      Page 29                                                GAO/NSLhD90-78       Navy Leadtime      Forecasts
                     Appendix I
                     Comments From the Department        of Defense




              procedures,e.g.,      restrictions        on unpriced    orders   (see discussion   in
              Finding    D), the Competition         in Contracting     Act, and the Refund Clause
              issue had significant          impact on the contracting        environment    and had a
              tendency to increase         administrative       lead time.

              There are two mechanisms which affect              in-file    administrative         lead
              time forecasts.        First,     forecasts    are updated on an item-by-item
              basis,    as contract     deliveries       are made and observations          of
              administrative      lead time become known.            The routine      transactions      can
              be thought     of as adjustments         to an existing     baseline--the        update
              mechanism has been in place for several                decades.

              The second update mechanism may be thought                     of as a baseline
              adjustment.        From time to time, the Aviation                Supply Office       becomes
              aware of major business              changes on the horizon          that can be expected
              to have significant            impact on its administrative              lead time.     TO
              properly     anticipate        these changes, in-file          administrative       lead time
              forecasts      are adjusted        in a programmatic        way. The particular           nine
              month adjustment          cited in Finding       D, below, is an example of such a
              non-routine       change.        The contention      that "they (the Aviation            Supply
              Office)     grouped items because           administrative        lead time determinations
              for individual        items would require          input from inventory          managers on
              each item and they did not want to do that"                      no longer applies.
Seecommentl   Administrative        lead     time forecasts      are routinely         updated on an
              item-by-item       basis,      using actual     administrative         lead time
              observations.

              FINDING C: Administrative            Leadtimes For Individual        Items Deviate
              Substantiallv        From Standards.       The GAO analysis     of 2,467 items
              purchased      during    a recent five month period        showed that the actual
              administrative        lead time for 863 purchases        varied   from the standards
              by at least six months.           In other words, the GAO found that--in           35
              percent     of the purchases      analyzed--the    Aviation     Supply Office either
              overstated      or understated      administrative    lead time requirements       by
              six months or more.

              The GAO also randomly sampled 1.50 stock items.                   However, the GAO was
              able to obtain       requirements      data on only 21 items because of the
              inaccuracies      in the Aviation       Supply Office      files.    The GAO review of
              these 21 items showed that the number of days used in the
              administrative       lead time requirements          computations    were overstated
              for 15 items ($645,000)           and understated      for six items ($796,000).
              The GAO cited examples were (1) canopy glass assemblies                    for A-6
              aircraft,      (2) digital    microcircuits       for the F-14 aircraft,       and



                                                               4




                     Page30                                                GAO/NSLUWO-78       Navy   Leadtime   Forecasts
                                 Appendix   I
                                 Comments       From   the Department   of Defense




                           (3) breather       adapters     for the A-6.      The GAO again concluded        that                       1
                          administrative         lead times used in requirement          determinations       should
                          be based on actual           experience     for individual   items.     The GAO further
                          concluded      that,     to the extent that representative,           Past experience      is
                          used to develop current             requirements,     those requirements      will provide
Now an pp. 3,9, 12, 13,   a more accurate          and realistic      basis for procurement      decision.
and15                      (p. 4, p. 13, pp. 18-21, p. 23/GAO Draft Report)

                          poll Response:         ConcUr.       The Aviation      Supply Office     has focused
                          command attention           on improving       the business process of contract          award
                          as part of its Total Quality                Management efforts.         The current
                          approach is to use past experience                  as a benchmark from which to
                          achieve      future     improvements.         This policy     is represented    by the
                          current      setting     for the administrative           lead time filter     parameters,
                          which are 0.01 percent              and 110 percent.         That choice of parameters
                          liberally        accepts declining        observations       and is designed     to ensure
                          their     recognition,         while limiting      the likelihood      of recognizing     an
                          aberrant       increase.

                          Another    current     initiative      improves the timeliness       of administrative
                          lead time forecast           updates.    Normal Navy Uniform Inventory           Control
                          Point procedures          record observations       for both administrative          and
                          production      lead times at the same time, when contract               deliveries
                          occur I The Aviation            Supply Office    has implemented    procedures        to
                          capture    and record administrative           lead time observations         shortly
                          after    contract     award, which is approximately          one production        lead time
                          earlier    than would otherwise          be the case.      As discussed     in the DOD
                          response to Finding            B, lead time calculations      are now done on an
                          item-by-item       basis.

                          The past several       years (particularly      in 1984 to 1987, the vintage          of
                          the data used in this audit)          have seen significant      changes affecting
                          the ability    of Government agencies         to award contracts     expeditiously.
                          Restrictions    on unpriced    orders,     the Competition    in Contracting        Act,
                          and the Refund Clause are just three examples of contributors                    to
                          increased    administrative     lead time.

                          FINDING D: Administrative               Leadtime Additive        Increased   Purchases.
                          The GAO found that,           during    one period     in 1986, rhe Aviation        Supply
                          Office    arbitrarily       added nine months to the administrative               lead time
                          standard     for all purchases          simply because funds were available.               The
                          GAO concluded         that,   by using the additive,          the Aviation    Supply :>ffice
                          raised the reorder          levels--thereby        generating      purchases  that -
                          otherwise      would not have been made or would have been made for lesser



                                                                              5




                                  Page31                                              GAO/NSIAIHW-78     Navy Leadtime     Forecasts
                                  Appendix I
                                  Comments From the Department          of Defense




                          quantities.           Examples of increased            purchases     cited by the GAO
                          involved         (1) direct      current     motors for the H-3 helicopter              and (2)
                          toggle        switches     for the H-l.         (The GAO did note that the additive                   was
                          a one time occurrence--which                  was eliminated       after the purchases            it
                          generated         were initiated.)           Since requirements         data on 64 consumable
                          items affected            by this additive        were available        from a prior        review of
                          operating         levels,     the GAO used the data to determine               the effect         of
                          the additive          on the Aviation         Supply Office purchases.             The GAO
                          compared the actual purchase values for the 64 items with what the
                          values would have been without                   the additive--revealing           that the total
                          purchase value would have been reduced by $2.1 million                              (from
                           $10.6 million          to $8.5 million).           According     to the GAO, by eliminating
                          the additive,           50 purchases        would have been made for lesser                quantities
                           and 14 purchases            would not have been made at all.                The GAO concluded
                           that the use of additives                to increase      administrative      lead times
                           artificially          should be limited,          reasonable     and fully     justified.           The
                          GAO further          concluded       that the availability          of extra funds is not
Nowon pp.3,14,   and 15    sufficient         reason to increase          requirements      in order to prematurely
                           purchase additional              stocks.      (p. 5, pp. 21-23/CiAO Draft Report)

                          Doll Response:       Nonconcur.       In FY 1986, the Aviation              Supply Office was
                          tasked by the Assistant           Secretary      of the Navy (Shipbuilding                and
                          Logistics)     to reduce the number of unpriced                 Basic Ordering          Agreement
                          actions     by 20 percent      and to reduce the dollar             value of unpriced
                          orders by 30 percent.            Since 70 percent          of the Aviation         Supply Office
                          obligations      had historically        consisted        of unpriced      orders,      compliance
                          with the firm fixed price policy                would have resulted            in the extension
                          of administrative        lead time by as much as 270 days for every sole
                          source buy.       As a consequence         of this acquisition           change, and not, as
                          the audit states,        "Simply because funds were available,"                      a one-time
                          Supply Demand Review was initiated                 with an additional            administrative
                          lead time.       This one-time       nine month administrative               lead time
                          additive     was to preclude        a potential        stockout    position       during     the
                          transition     from an unpriced         order environment          to a priced         order
                          environment.        The impact on Stock Fund expenditures                    resulting         from
                          this lead time additive           was limited        to increased       safety level
                          requirements      due to increased         variability        in lead time.

                          The Aviation        Supply Office         has ensured and will            continue     to ensure
                          that arbitrary        requirements         are not procured.

                          FINDING E: <.                                                                                  The GAO
                          reported   that,  unlike          administrative      lead times, the Aviation                 S~~pply
                          Office   does determine           production     lead times on an item-by-item                  basis,



                                                                               6




                                   Page 32                                                  GAO/NSLAD-90-78       Navy Leadtime       Forecasts
                                    Appendix I
                                    Comments From the Department       of Defense




                            The GAO observed that,             in order to make realistic            forecasts      and
                            ensure that neither            too many nor too few spare parts are purchased,
                            DOD policy        states that production         lead times      used in the requirements
                            should be based on estimates              from contractors         or historical
                            information         that has been collected         for representative           procurements.
                            In line with this policy,              the GAO found that the Aviation               Supply
                            Office      collects      data on actual     experienced      production       lead times and
                            contractor        estimates.        The GAO further     found, however, that the
                            production        lead times used in requirement             computations        often do not
                            realistically          reflect    the time actually       required     to receive materials
                            from contractors.

                            The GAO reviewed the production                   lead time requirements      for 21
                            consumable      items and found that the lead times used in the
                            requirements       computation        differed       from the actual   production     lead
                            time in every case.            According        to the GAO, production      lead times for
                            17 items were overstated              by up to 472 days and the production              lead
                            times for four items were understated                    by up to 631 days.        The GAO
                            substituted       the actual production              lead times for those used in the
                            requirement       computations        and determined        that, on total    purchases
                            of $7.4 million,         requirements          for the 17 items were overstated           by
                            $1.6 million        and requirements           for the four times were understated           by
                            $43,000.       The GAO concluded           that, as with administrative           lead time,
                             inaccurate     production        lead time can have a significant            adverse
Now on pp, 4, 17, 18, 22,   effect--either         generating       excess stocks or causing shortages.
and 23                        (pp. 4-5, pp. 25-28, pp. 35-36/GAO Draft Report)

                            POD Response:           Concur.       Production     lead time forecasts             indeed have a
                            degree of inaccuracy,             as do any forecasts.             However, the key issue is
                            whether a better          forecast       could have been provided,              given the data
                            available       at the time of the forecast.                  The fact that the forecasts
                            differed      from the actual observed values is not surprising.                              The
                            degree of variation           is a valid        concern.       The analysis        in the audit
                            should be extended.             The GAO concludes           that the reason the variations
                            were so large was because the Navy was not using the GAO-recommended
                            forecast      system,     which is to base the next production                     lead time on
                            the last one.           The Navy uses this approach,               but with some
                            modifications         to take into account:              (a) how long it has been since
                            the last procurement             (i.e.,     if the last contract           was five years ago,
                            is it truly        representative         of today’s      procurement        lead time?) and
                             (b) the degree of variation                of the most recent observation                 from the
                            historic      norm.      The Aviation         Supply Office      is cited not for the
                            system, but the degree of difference                     (called    filters)       allowed between
                            past history        and recent observations.                The Aviation        Supply Office had



                                                                              7




                                    Page 33                                               GAO/NSL4D-90-78      Navy Leadtime      Forecasts
         Appendix I
         Comments From the Department      of Defense




been using 80 percent         and 125 percent      as filters,      meaning that a lead
time observation        less than 80 percent,        or more than 125 percent,        or
the historic      lead time, would not be used to update that lead time
forecast.      Due in part to the preliminary            audit results,     the Aviation
Supply Office       evaluated    its forecast    filters      and changed them to
reflect    better    values for updating      the historic       forecasts.

The parameters     now in use are 70 percent    and 110 percent;      this change
is   intended  to record the downward trend in production        lead time that
the Aviation     Supply Office   is working to achieve.      In addition,   Navy
has a pro-active      initiative   underway to review data to determine
whether a further       lowering of the 70 percent   parameter   is
appropriate.

Forecasts      will     always have a degree of inaccuracy                in them; if not,
they would not be forecasts               as we know them.         Technology      in the areas
of major investment--high              technology       weapon systems--is        changing    so
rapidly     that it affects         the ability       to accurately      predict     item
demand, procurement           lead time, and other factors              that must be
considered.          The operating       environment,      number of weapon systems in
the program,         or even the state of world alliances                 cannot be predicted
with unfailing          accuracy;     all of these will        also lead to greater          or
 lesser degrees of accuracy in the forecasting                      systems.      The Navy is
actively      pursuing      a number of initiatives           designed      to provide    better
procurement         decisions     as to quantity        and timing of orders and will
continue      to try to improve the process.

Findina    F:      Lead Times That Exceed Parameters           Are Not Reviewed.           The
GAO learned        that,     in determining    and forecasting      production      lead time
requirements,          the Aviation     Supply Office    automatically       screened out
experienced        lead times that are not within           80 to 125 percent         of the
leadtime     on    file.       The GAO observed that these parameters            are
intended     to    ensure that atypical        data do not influence         the
requirements          forecast.

The GAO found, however, that the Aviation               Supply Office does not
review the lead times that are outside              of the parameters     to determine
if they are actually          representative     as opposed to atypical.        Of the
21 consumable       items   it reviewed,     the GAO found that 14 items were
outside    the parameters.          The GAO cited,   as an example, the case of
compressor    blades,      where the Aviation      Supply Office    files   continued
to show a production          lead time of 573 days because the latest
experienced      production      lead time of 147 days and the two prior
production     lead times of 200 and 349 days were below the 80 percent



                                                 8




         Page 34                                             GAO/NSLAD-90-78     Navy Leadtime     Forecasts
                                   Appendix I
                                   Comments From the Department        of Defense




                            parameter       and, therefore,        were not reviewed.       The GAO concluded       that,
                            in the case of the compressor blades,                 this contributed    to a
                            subsequent       excess of blades.          The GAO pointed     out that, while DOD
                            policy     states that when experienced           production     lead times Or
                            contractor       estimates      are determined    to be nonrepresentative         of future
                            performance        they should be excluded        from normal production        lead time
                            development,         the policy     also emphasizes that these data should be
                            retained      in the procurement         history  file.      The GAO concluded      that,
                            unless production          lead times outside       the parameters     are reviewed,       the
                            Aviation      Supply Office        has no way of knowing if they are
Now on pp. 4, 18, 19, 22,    representative         or nonrepresentative.          (pp. 6-7, pp. 28-29,
and 23.                     pp. 35-36/GAO Draft Report)

                            DoD Resuonse:          Concur.      Current     Uniform Inventory         Control     Point
                            computer system design requires                 the selection       of two parameters,
See comment   2.            upper and lower production              lead time filter         parameters,       which are
                            used to filter         out "atypical"        production      lead time observations.
                            Observations        that pass through          the filter      automatically       update the
                            production       lead time estimate          in file.      As a matter of policy,
                            observations        that fail     to pass through         the filter      may either      be held
                            in suspense pending manual review or may be discarded                          from the
                            production       lead time observation             update process.        It should be noted
                            that,    contrary      to the statement          in the report,       all observations       are
                            recorded     in the contractor          status file as part of the procurement
                            history,      regardless      of whether or not they are used to update the
                            in-file     production      lead time estimate.           In addition,       it should be
                            recognized       that the resource          expenditure,       which would be required          to
                             review all re jetted         observations,         would be substantial          and, in the
                            DOD view, unjustifiable.

                            The choice of parameter          values is used to balance the risk of
                            automatic     file    update with inventory      manager workload.         During the
                            1986 command inspection          of the Aviation    Supply Office,       it was noted
                            that with the parameter          values then in effect,       approximately      2,000
                            observations       per week were released       for inventory      manager manual
                            review and, of these, approximately             90 percent    were subsequently
                            deleted    from further      consideration.      On that basis,      the decision      was
                            made to cease manual review of such observations                 in order to improve
                            inventory     manager productivity.

                            The current    choice of parameters    (70 and 110 percent)   is intended                        to
                            maintain    the existing level of productivity    while recording     the
                            downward trend on production      lead time that the Aviation     Supply
                            Office   is working to achieve.      A number of programs have been



                                                                              9




                                   Page36                                                 GAO/NSIAIHO-78       Navy Im&ime        lbn?c~
        Appendix1
        CoxnmentsFromtheDepartmentofDefense




developed,      including:           (1) in accordance         with DOD initiatives.               the
Aviation     Supply Office          procurement      specialists        are soliciting
contractor      deliveries         25 percent     earlier      than file       data    would
otherwise     indicate;         (2) increased      use of indefinite             delivery       type
contracts;      and (3) implementation             of acquisition          planning
initiatives,       which will         lead to a general          streamlining         of the
procurement       process.         For example, there are procurement                    automation
via the Procurement             Early Development         System, improved tracking                  of
procurement       referrals        via an automated         system, improved support                  from
the Naval Air Technical               Support Facility         for procurement           drawings,
and better       support      from Navy printing          offices      for procurement
printing,     quicker       proposal      turnaround      time using computer
communications         media, and better          document distribution               through       the
development       of the Solicitation            Process Automation            project       with Navy
printing     offices.

The DOD recognizes           that the choice of production                 lead time filter
parameter       values has important             ramifications        for many aspects of the
Inventory       Control     Point business,          including      requirements
determination,          item management workload,               and procurement        workload.
When properly         selected,        they provide       an economical       means of assuring
that automatic          production        lead time file        updates are, with a high
level of confidence,              representative        of "typical"       conditions,      while
simultaneously          assuring       that rejected        observations      are, with a high
level     of confidence,          representative        of "atypical"        observations.
Current     and future        initiatives        to improve the choice of filter
parameter       values offer          a more practical         and cost-effective        method of
increasing       the accuracy of lead time projections                     than does the
labor-intensive           manual review of every rejected                 observation.

FINDING G: Greater           Use Can Be Made of Contractor               Estimates.
The GAO also reported           that,     in addition    to historical         information
on experienced        lead times, contractor          estimates       are to be used in
forecasting       production      lead time requirements.             The GAO found,
however,     that the Aviation          Supply Office      generally      does not obtain
contractor      estimates     before determining         production       lead time
requirements.         The GAO further        found that,      in those instances           when
estimates      are received,        they are not compared with actual
performance.         The GAO pointed        out that DOD policy          states that such
estimates      and comparisons         are necessary     so that the most accurate
production       lead time forecasts         can be made.        The GAO found, however,
that the Aviation          Supply Office       (1) obtains      and uses production          lead
time estimates        from only five major contracts,                (2) does not maintain
statistics      on the number of individual            contract       estimates      received,



                                                      10




         Page36                                                     GAO/NSIAIWO-78NavyLeadtimeForecasts
                                    Appendix1
                                    Comments    From the Department       of Defense




                            and (3) estimates   that contractor                information       is applied      only    to
                            about 5 percent   of the items in               inventory.

                            The GAO concluded        that,   if contractor     estimates     were routinely
                            obtained    and used in calculating         production      lead time requirements,
                            more accurate       forecasts    would result.       The GAO also concluded      actual
                            contractor     performance      should be compared with the contractor
                            estimates     and the production       lead time used in the requirements
                            computation     to assure future       production      lead time requirements     would
                            be more accurate.           (The GAO noted, however, that unfortunately           the
                            Aviation    Supply Offices       does not have a procedure          for making such
                            comparisons.)

                            The GAO observed that positive            action      has been directed     in the area
                            of contract    negotiations.        The GAO reported         that the DOD has been
                            concerned   about the length of lead times and, in May 1989, held a
                            major conference      on reducing     lead times.         The GAO reported       that,
                            while in the past Navy procurement               policy   was not clearly     oriented
                            toward negotiating        reduction    in procurement        lead times, in May 1989,
                            the Secretary     of the Navy reported           to the DOD that this policy           had
                            changed and that the Navy was now negotiating                   procurement    lead time
Now on pp. 4, 19, 20, 22,   reductions    with contractors.          (pp. 6-7, pp. 30-31, pp. 35-36/GAO
and 23.                     Draft Report)

                            DOD Response:          Partially        concur.      The DOD agrees that information
                            from contractors           is valuable       in the procurement           process for a number
                            of reasons,       including        better    estimates      of production         lead times.        The
                            DOD does not agree, however, with the GAO interpretation                                of DOD
                            policy,     as set forth in DOD Instruction                   4140.55.       Specifically,        the
                            GAO statement        that "DOD policy           states that such estimates                and
                            comparisons       are necessary           so that the most accurate             Production       Lead
                            Time forecasts         can be made" conflicts              with the statement           in the
                            policy    that "Production            Lead Time may be based on estimates                    from
                            contractors,        historical        information       that has been collected              for
                            representative         procurements,          or provisioning         technical
                            documentation."             Inventory      managers,      as part of the requirements
                            determination         process,      validate      production       lead times using a number
                            of data sources,           which include        historical        data and contractor
                            furnished      estimates,        when available.           Recognizing       that contractor
                            furnished      data can be unreliable               (as the GAO noted in the area of
                            cost estimates          in GAO/NSIAD-88-7,           "Contractor        Cost Estimating
                            Systems,"      OSD Case 7538) and that the contractor                       suffers     no penalties
                            for inaccurate          estimates,        such data are but one source of data among
                            several     used in the validation              process.        It should be noted that the



                                                                                11




                                    Page37                                                    GAO/NSIAD-90-78      Navy Leadtime       Forecasts
                                     Appendix I
                                     Comments From the Department           of Defense




                                                                                                                                               -

                            Aviation      Supply Office        obtains    contract      administrative       data
                             (including      expected contractor          deliveries       based on performance
                            monitoring)        via access to a realtime             Defense Logistics        Agency
                            contract:     administrative        data base.         The Defense Logistics          Agency
                            currently       administers       66 percent      of the Aviation         Supply Office
                            contracts       and it is anticipated           that, by the end of FY 1990, the
                            Defense Logistics          Agency will administer             all  of the Aviation        Supply
                            Office      contracts.       This will     give the Aviation          Supply Office       an
                            increased       capability      to validate       projections      of production       lead time.
                            The use of such real time data is a more reliable                          and appropriate
                            source of information             on contractor        performance      than contract
                            estimates.

                            FINDING H: Historical                  Records Are Not Complete:                 Internal       Controls
                            Need To Be Improved.                  The GAO reported            that the records          used in the
                            requirements         determination           process contain           numerous inaccuracies.
                            For example, a GAO analysis                    of 150 randomly selected               items showed
                            that shipment and receipt                   records      in the contract         status file did not
                            agree for 122 of the items (or about SO percent).                                  The GAO projected
                            that about 4,200 items, involving                        purchases      of $487.5 million             (report
                            table 3.1) had file               discrepancies          between shipments and receipts.
                            The GAO reported             that,      according      to Aviation        Supply Office         officials,
                            a major reason why accurate                    records axe not maintained                 is that the
                            data files       are fed information                from a variety          of sources and this
                            leads to errors,             mismatches,         nonpostings,         and other problems.              The
                            GAO observed         that the discrepancies                  within    and between the Aviation
                            Supply Office          files       indicate      that internal         controls     over file        data
                            need to be improved.                  The GAO also noted that its review of annual
                            assurance      reports         indicated       the Navy has not identified                  inaccuracies
                            in the file       data as a significant                  weakness.        The GAO concluded            that
                            incomplete       and inaccurate             inventory         records    further    hamper lead time
                            forecasting.           The GAO also concluded                  that, without       reliable       data, the
                            Aviation     Supply Office              does not have reasonable               assurance that the
                            procurement        system is adequately                protected       from waste, fraud,            and
                            abuse.      Finally,         the GAO concluded              that more accurate          requirement
N W on pp. 4,20,21,   22,   determinations           would reduce the chances of generating                         excess stocks
and 23.                     because of overstated                 procurement         lead time requirements.                (PP. 7-8,
                            31-35, p. 36/GAO Draft Report)

                            DOD Response:      Concur.   Resystemization     is a major Project     to
                            redesign    completely   the software   system that supports      all the
                            functions    of the Navy Inventory     Control   Points.   The project     is
                            divided   into four major phases, with 90 percent         of the effort
                            occurring    in Phases II, III,     and IV.    The next scheduled




                                     Page 38                                                     GAO/NSIAD-St%78       Navy Leadtime        Forecasts
--        _
                                Appendix   I
                                Comments       From the Department   of Defense




                       R&systemization          release     is Phase II (the delivery             schedule is
                       currently      in jeopardy        due to technical          problems and resource
                       limitations),         which includes        financial       and supply material          tracking
                       upgrades.        Inherent      in the new design is elimination                of duplicate
                       supply and financial             data bases, and merger of inventory                 and
                       financial      systems into one overall               integrated     file   structure.
                       Implementation          of the Uniform Chart of Accounts material                    accounting
                       system will        ensure that assets procured                and paid for by the Navy are
                       properly      reflected       on all accountable           records   after DD 250 signature
                       acceptance       by a Government representative.                   The current      1960's
                       systems design has long been recognized                       as an impediment       to accurate
                        lead time forecasting.               That is part of the basis for the significant
                       Navy resource         commitment involved           in the Resystemization            effort.      The
                       Navy has, in the interim,               established        a flag level      Inventory       Accuracy
                        effort     to establish        as many improvements           as possible      regarding
                        shipments      and receipts.

                       As cited      in the DOD response to Recommendation     7, data accuracy
                       problems      will  be targeted   as an issue for review in FY 1991 under                         the
                       Financial       Manager Integrity   Act assessments.



                                                                     RECOMMENDATIONS

                       pECOMMENDhTION 1: The GAO recommended                that the Secretary       of the Navy
                       direct    the Commander, Naval Supply Systems Command, to implement
                       procedures      that more accurately       determine     administrative     lead time
                       requirements--by        basing administrative        lead time requirements        on
                       representative       actual   experience    for individual        items rather than
Now ori pp, 5 and 15   standard     lead times for groups of items.             (p. 8, p. 23/GAO Draft
                       Report)

                       DOD Response:        Concur.      The recommendation        is moot, however, inasmuch
See comment   1,       as administrative         lead time requirements         are, in fact, based on
                       representative       active    experience     for individual      items and not on
                       standard     lead times for groups of items.               (The response to Finding  B
                       contains     further    discussion     of this point.)

                       RECOMMENDATION 2: The GAO recommended that the Secretary            of the Navy
                       direct      the Commander, Naval Supply Systems Command, to implement
                       procedures       that more accurately   determine   administrative  lead time
                       requirements--by       ensuring   that additives  to administrative    lead time



                                                                          13




                                Page 39                                                 GAO/NSLAD-90-78      Navy Leadtime      Forecasts
                             Appendix I
                             Comments From the Department       of Defense




Nowon     p. 15       requirements      are   reasonable     and fully       justified.        (p. 23/GAO Draft
                      Report)

                      DOD Response:      Partial1    y concur.    The DOD agrees that                additives  to
                      requirements     should be reasonable       and fully    justified,             but contends
                      that appropriate       safeguards    are already    in place.

                      Recommendation      3: The GAO recommended that the Secretary                 of the Navy
                      direct   the Commander, Naval Supply Systems Command, to implement
                      procedures     that more accurately       determine    administrative       lead time
                      requirements--by      reviewing   historical      data on administrative         lead
Nowonp.     15        time requirements       for completeness      and accuracy.         (pp. 23-24/GAO
                      Draft Report)

                      DOD Response:         Concur.      The Aviation      Supply Office       has    implemented
                      procedures      to more accurately         determine    administrative          lead time
                      requirements.         Historical      data are already       included     in    the
                      maintenance       of administrative        lead time.      In November         1989, the
                      Aviation    Supply Office         reset the procurement         lead time        filters   to
                      ensure inclusion         of all internal       process improvements            affecting
                      contract    award.       To the extent that one-time            aberrant       experiences    are
                      eliminated      by these filters,         unnecessary     manual review          in a limited
                      personnel      resource     environment      has been avoided,         The     cited practice
                      is consistent       with DOD direction         to reduce inventories             and lead
                      times.     In addition,        the current     data review cited in            the DOD
                      response to Finding           E is aimed at determining           whether      a further
                      decrease     in the 70 percent         parameter     is appropriate.

                      Recommendation        4: The GAO recommended that the Secretary               of the Navy
                      direct    the Comander,       Naval Supply Systems Command, to implement
                      procedures      to ensure that more accurate           production     lead time
                      forecasting       data are available      and used in making requirement
                      determinations--by        reviewing    experienced       lead times and contractor
                      estimates     that are outside       of established        parameters    to determine  if
Now on pp. 5 and 23   they are representative          of current     conditions.        (p. 8, pp. 36-37/GAo
                      Draft Report)

                      DOD Response:   Partially  concur.     The DOD does not agree with the
                      recommended use of contractor     estimates.   (The issue is addressed                         in
                      detail  in the response to Finding     G.)
See comment       2   The DOD concurs that the assurance                 that experienced    lead times are
                      representative  of current   conditions               is vital  to the effectiveness
                      and efficiency  of Inventory   Control              Point operations.      The DOD



                                                                      14




                             Page 40                                                  GAO/NSIAD-90-78   Navy Leadtime     Forecasts
                          Appendix I
                          Comments From the Department          of Defense




                   contends,   however, that this assurance           is more economically     obtained
                   from the careful     selection     and periodic      analytical    review of the lead
                   time filter    parameter     values rather     than through     the manual reviews
                   and processing     of individual      update transactions.

                   Recommendation        5: The GAO recommended that the Secretary           of the Navy
                   direct     the Commander, Naval Supply Systems Command, to implement
                   procedures      to ensure that more accurate        production    lead time
                   forecasting       data are available       and used in making requirement
                   determinations--by        routinely     obtaining  contractor   estimates   of
                   production      lead time and comparing the estimates          with actual
Nowon     ~1.23.   contractor      performance.       (pp. 36-37/GAO Draft Report)

                   DOD Response:        Nonconcur.      As stated in the DOD response to
                   Finding     G, production     lead time may be based on several                 sources of
                   information,      of which contractor          estimates     is but one.        Contractor
                   estimates     can be unreliable       and the contractor           suffers     no penalty    for
                   unreliable     estimates    provided     prior     to contract       award.     The relevant
                   information      used by the Aviation          Supply Office        (real time data from
                   the Defense Logistics        Agency contract          administration        data base) is a
                   more reliable       and appropriate      source of information             on contractor
                   performance.

                   Recommendation         6: The GAO recommended that the Secretary             of the Navy
                   direct     the Commander, Naval Supply Systems Command, to implement
                   procedures       to ensure that more accurate         production     lead time
                   forecasting        data is available       and used in making requirement
                   determinations--by           improving   the completeness       and accuracy of
                   production       lead time data in the historical           files   and establishing
                   the internal        controls     needed to ensure the accuracy of these files,
                   including      the reconciliation        of shipment and receipt        data in the
Nowonp.      23.   contract      status file.         (pp. 36-37/GAO Draft Report)
                   DOD Response:           Partially       concur.       Improvement in completeness                  and
                   accuracy of production              lead time data in historical                 files      (i.e.,
                   contract     status file)         is a desirable          goal.     As cited in the DOD
                   response to Recommendation                 7, data accuracy problems will be targeted
                   as an issue for review in FY 1991 under the Financial                                Manager
                   Integrity     Act assessments.               In terms of its contribution                 toward
                   improving     accuracy of production                lead time forecasting              data, however,
                   this goal is of secondary                importance       relative     to the several            other
                   initiatives      discussed        elsewhere       in the DOD response.               The numerous
                   contracting       initiatives         targeted      at reducing       production        lead times
                   are critical        to success in improving               forecasting      accuracy.           Careful



                                                                      15




                          Page 41                                                   GAO/NSLAB90-78       Navy Leadtime      Forecasts
                     Appendix I
                     Comments From the Department          of Defense




              selection     of production        lead time forecast       filter parameters      (i.e.,
              real time contract          administration        data bases and, to a lesser extent,
              contractor-provided          delivery      estimates)   are both highly    significant
              forecasting      process improvements.             It is in these areas that the Navy
              is currently       focusing     its efforts.

              Resystemization,            the Navy's ongoing effort             to modernize        Inventory
              Control      Point data processing             systems, will provide            substantial
              mechanized capabilities              needed to better          manage lead time
              forecasting.         Phase II of this project               (the delivery         schedule     is
              currently       in jeopardy        due to technical         problems and resource
              limitations),        will      combine currently         independent        supply,     procurement,
              and financial         files     into a single       integrated        data base.        A
              fundamental       feature       of the new design is that both its item lead time
              observations        and contract         record history        will be concurrently            updated
              upon processing           of material       receipt    reports      from procurement
              consignees.         The result        should be considerable             reduction,       if not
              elimination,        of further        disconnects      between supply data and contract
              records.

              The current     1960s system design has long been recognized                    as an
              impediment    to accurate       lead time forecasting            and that is part of the
              basis for the significant             Navy resource     commitment involved        in the
              Resystemization       effort.      As an interim       measure, the Aviation         Supply
              Office    has established       a Process Action Team to examine possible
              short term workarounds          until    the Resystemization          effort  is complete.
              Short term efforts         to alleviate       the effects       of these system
              weaknesses include         the design of an Obligation              Status File/Contract
              Status File/Due-In         Due-Out File diagnostic             (due to be completed by
              March 1990) and implementation               of direct    access to the Defense
              Logistics    Agency contract          administration      file.

              Recommendation      7: The GAO recommended the Secretary         of the Navy
              direct    the Commander, Naval Supply Command, to implement procedures
              to ensure that more accurate       production    lead time forecasting    data
              are available      and used in making requirement     determinations--by
              targeting     data accuracy problems as an issue for review in the
Now on p 23   Federal Manager‘s      Financial  Integrity   Act assessments.
               (pp. 36-37/GAO Draft Report)

              m:                     Concur.  Data accuracy problems                 will be targeted         as an
              issue for review         in FY 1991 under the Financial                 Manager Integrity         Act
              assessments,




                                                                 16




                      Page 42                                                  GAO,‘NSlAD-90-78     Navy Leadtime      Forecasts
--   -
               Appendix I
               Comments From the Department   of Defense




               The following    are GAO'S comments on DOD'S letter dated April         17, 1990


GAO Comments   However, when standard ALTS are placed in computer files they auto-
               matically override collected experiences and result in minimum or maxi-
               mum AI,TS that inventory managers are required to use in making
               requirement determinations.

               2. Although  we agree that the parameters can isolate apparent atypical
               PLTS, we continue  to believe that PLTS outside of the parameters should
               be reviewed to determine if they are truly nonrepresentative.    Our tests
               showed that a large percentage of PLTS was outside of the parameters.
               The Navy has about $1.6 billion in PLT requirements for consumable avi-
               ation parts, We believe that reviewing screened out PLTs can improve the
               accuracy of PLT forecast? and minimize the chances of buying excess
               material or having stock shortages. Even if only 10 percent of the fore-
               casts were revised, $160 million in PLT requirements could be potentially
               affected.




               Page 43                                     GAO/NSIAD-90-78   Navy Leadtime   Forecasts
Appendix    II

Major Contributors to This Report


                            James Murphy,   Assistant Director, Navy Issues
National Security and
International Affairs                                                                                            Y


Division,
Washington, D.C.

Philadelphia     Regional   Edward Fossler, Evaluator-in-Charge
Office                      George Surosky, Site Senior
                            Keith Steck, Evaluator
                            Thomas Bloom, Systems Analyst




 (394277)                   Page 44                                GAO/NSLAD-90-78   Navy Leadtime   Forecasts
          _’
                                                                                                                             . .._.;.          -.            ;
           .-                                                                                                                ,_”        .,,; .~;
                                                                                                                             _’         .-
                                                                                                                                      .. .
                                                                                                                                      ‘. :.


                                                                                                                       :         ..                      :
                :                                                                                                          .~(            ,.

                    :.
                                                                                                                                                    “’
                                                                                                                                 .:_




                                                                                                      .’




                                                                                              .,-    ._.   ,‘. 1_ ,;
,’   ,’                  ‘-   ‘:   ,(.   .;;   ..(   ‘   ..   .;   :   .,,.’   .   .,.’   ,    _.: