- _ United States General Accounting Ofllce / qC’20 -2 “s‘GAO Testimony 142202 For Release Competiti,veness of Federal on Delivery Computer Procurements Expected at 10:00 a.m. EDT Sept. 13, 1990 Statement of Milton J. Socolar, Special Assistant to the Comptroller General Before the Subcommittee on Legislation and National Security Committee on Government Operations House of Representatives Y GAO/T-IMTEC-90-12 GAO Form 160 (12/87) Mr. Chairman and Members of the Subcommittee: I am pleased to participate in the SubcommitteeVs hearings on the effectiveness of competition for automated data processing (ADP) procurements in the federal marketplace. Our work, which involved reviewing procurements at several agencies and compiling and analyzing governmentwide statistics on mainframe computer buys, provides a good starting point for assessing the issues surrounding the competitiveness of federal computer acquisitions. Our work on individual acquisitions has shown that agencies have many difficulties in properly acquiring computer hardware and software-- including three reports which show how government agencies inappropriately restricted competition. Our governmentwide statistics show that 82 percent of all mainframe procurements limit competition in some way. I should add, though, while that statistic does raise concern as to the extent of competition limited procurements, such limitations are not necessarily inappropriate or inconsistent with requirements of the Competition in Contracting Act. We have not examined the individual procurements covered by our governmentwide statistics to determine the extent to which restrictions on competition were unjustified. T OPaING ENVIRONMENT Competition is a critical factor in government procurement law and policy. Generally, the government is best served when all potential contractors have an equal opportunity to compete with others for its business. Offering all contractors the opportunity to compete helps to assure that the government pays fair and reasonable prices. When competition is unnecessarily restricted, the government may pay too much for what it buys and may not obtain the best solution for meeting its needs. The Brooks Act and the Competition in Contracting Act of 1984 (CICA) are the primary laws governing the federal acquisition of ADP equipment. These two laws and their implementing regulations mandate that agencies use full and open competition in conducting ADP procurements unless restrictions are appropriately justified. Brooks Act purchases must be based on full and open competition, using specifications that identify no more than an agency's minimum needs. The Act directs the Administrator of the General Services Administration (GSA) to provide for the economic and efficient purchase, lease, and maintenance of ADP equipment. To help meet this responsibility GSA delegates its procurement authority to individual agencies for certain procurements but maintains its overall oversight responsibilities. Under CICA, agencies are allowed to restrict specifications only as necessary to meet their minimum needs. CICA states that agencies may limit Y 2 the number of potential vendors only when there is but one responsible source that can satisfy the need, when the government would be seriously injured unless the number of potential vendors is limited, or in special circumstances such as those that involve a national emergency or a compromise of national security. Many agencies are looking to technology to meet rapidly escalating demands for increased productivity. Yet, as most observers agree, agencies have had great difficulty in harnessing this technology. Procurements often take many months or even years to complete: to shortcut this process, agencies skip important steps in defining their needs. This increases the risk of buying the wrong hardware and software and of unnecessarily limiting competition. AGENCYACTIONS UNNECESSARILY IT COMPETITION Our reports have documented severe problems that agencies have in conducting ADP acquisitions. In several reports we found a number of factors which contributed toward procurements that are not fully competitive: improper or inadequate review of available alternatives to meet agency needs: 'poor definition of requirements: lack of experience, training, or knowledge in proper procurement procedures by agency staff: taking unreasonable steps to minimize risk: and, in some cases, showing improper bias towards a particular vendor. Common to all these procurements was a lack i) 3 of agency oversight in assuring that adequate and appropriate competition was being fostered in the acquisition process. We also found a lack of adequate oversight by GSA. Some examples should help to illustrate the point. c ’ In August 1987, the Navy awarded a $26.7 million contract for mainframe computers to support its Standard Civilian Pay System (NAVSCIPS). We found that the Navy, in its haste to meet a self- imposed deadline, selected a data base management system, hardware, and the 10 sites where the system would run, without conducting appropriate studies and developing adequate support.1 Taken together, these decisions dictated specific hardware which limited competition to one vendor. Further, the Navy's disregard of good acquisition practices resulted in a system design that would not work. We recommended that the Navy halt further hardware and software orders and conduct a proper cost-benefit analysis for a full range of system alternatives. The Navy concurred and has stopped delivery of the computers. Further, it has determined that it can use existing equipment and needs only one of the 10 sites it originally designated. 1ADP Procuremen t* . Navv Imnronerlv Restricted Comnetition for Its Cm (GAO/IMTEC-89-61, June 21, 1989). 4 This year we reported on the Federal Aviation Administration's (FAA) $1.5 billion Computer Resources Nucleus project (CORN). This project is intended to meet the agency's general purpose data processing needs for 10 years and provide support to other Department of Transportation organizations. In May we recommended that the CORN procurement not be awarded because it had not been properly justified and planned.2 In June we also reported on the restricted specification issue.3 Originally, FAA planned to have full and open competition and encourage vendors to propose innovative approaches. to meeting its needs. FAA later decided to require a single architecture on the basis that it would reduce operational costs and provide a technical platform for an integrated data base. We found that a single architecture would not necessarily reduce costs or assure an integrated data base and that therefore, FM was unjustifiably limiting competition and restricting the range of solutions that vendors could offer. FAA is considering our recommendation that it not award the procurement. 2FAA Procurement. . Manor Data-Processina Contract Should Not Be Awarded (GAO/IMTEC-90-38, May 25, 1990). 3FAA Procurement: Comnetition for Maior Data-Processina Proiect Was Unjustifiably. Limite d (GAO/IMTEC-90-71, June 11, 1990). * 5 . 9PM Retirement S-tea At this Committeels request, we reviewed the Office of Personnel Management's (OPM) planned $54 million procurement for an automated retirement system.4 We found that the request for proposals (RFP) required vendors to meet highly restrictive experience requirements as a condition of bidding on the contract. While we agreed with OPM that experience should be a factor in contract award, we felt the experience factors could be considered as an element of the technical evaluation rather than being set out as an absolute condition. Using these requirements as a condition of bidding unnecessarily limited the range of vendors who might bid on the project. OPM agreed with our recommendation and is reissuing the RFP without limiting factors. Further, OPM determined a compatibility-limited solution was not required. Denartment of Defense (DOD) Inspector General ReDOrtS The DOD Inspector General's office has issued audit reports examining whether various Navy contracts had been unnecessarily restrictive of competition. In December 1989, the Inspector General reported that the Navy's attempt to buy a communications processor for the Navy Military Personnel Command was 4Retirement Svstem: Concerns Abo t PM ERS Auto ted Processina System Procurement (GA~/I~TEr",~-45, Aprma4, 1990). 6 unjustifiably biased toward IBM.5 In a more recent report, the Inspector General found that the Washington office of the Navy Regional Data Automation Center did not use required competitive procedures for procuring communications equipment for the center.6 The Center required a specific make and model of equipment and did not advertise the delivery order in the Commerce as required to promote competition. ION FOR MAINFRAMECOMPUTERS IS INFREQUENT FEDERAL M.&R&ETPLACE At.your request, Mr. Chairman, we developed statistics on agency mainframe computer purchases totaling $1.9 billion for the 3-l/2 years ending in March 1989, including information on vendors and types of procurement action for the 35 largest federal agencies.7 During this past year we have issued individual reports on eight of these agencies (see attachment II for the titles of these reports). We plan to issue reports on two additional agencies and an overall capping report (see attachment I for charts depicting . 5yJaval Milltarv Personnel Command Planned Procurement of Automated Data . Processina . nt (Inspector ECtulDme General Report 90- 019, Dec. 15, 1989). . 6Navv Reaional Data Auto mation Center, Washinaton. D.C.. . nt (Inspector Procurement of Automatic D-i;a Processina Eculwme General Report 90-103, AugC( 24, 1990). 7The $1.9 billion represents obligated dollars for the 3-l/2 year period. The total value of the mainframe computer contracts exceeds $1.9 billion since some contracts extend beyond the 3-l/2 year period. governmentwide statistics). These statistics show that the government is limiting competition to acquire its mainframe and mainframe peripherals. Over 80 percent of the procurements in our review (total dollar value) required compatibility with existing systems. IBM-compatible procurements comprised 64 percent of the compatibility-limited procurements and 52 percent of all procurements. IBM was the single largest vendor with a 47 percent share of the total federal market. Unisys was the next largest with a 20 percent share. Almost 60 percent of the total value of all procurements involved only one vendor. Specifically, Modifications to Existing Contracts was the most popular procurement method used, representing 44 percent. New Contract-Sole Source and New Contract-One Offeror procurements represented 10 percent and 5 percent, respectively. Although these statistics describe what has occurred in the federal mainframe computer marketplace, they do not explain whether or not agencies have made smart acquisition decisions or have followed required regulations and laws in conducting their procurements. When coupled, however, with our work on individual procurements, the picture that emerges reflects a marketplace 8 dominated by competitively limited--and sometimes poorly justified--procurements. Two important issues arise from this situation. -- Agencies may be limiting competition for legitimate reasons. For example, some agencies have a large hardware/software base that effectively mandates compatibility. On the other hand, we know that a general lack of skills, knowledge, and understanding of ADP systems and the procurement process, coupled with ineffective agency oversight, can also result in inappropriate limitations on competition. Similarly, we found examples where competition was limited by a desire to expedite the procurement process or deliberately favor a particular vendor. -- GSA did not adequately exercise its oversight responsibility in our examples cited above. This raises the question as to how well GSA is discharging its responsibility for oversight of federal ADP acquisitions on an overall basis. Agencies are trying to cope with a rapidly changing world of technology. The issues raised above call for: -- Agencies to strengthen their control and oversight over ADP procurements; and for J -- GSA to take more of a leadership role (1) in assessing whether the current situation (i.e., numerous competition- limited acquisitions) is justified; (2) in working more effectively with agencies to assure that procurements comply with existing regulations: and (3) in ensuring that the government's methods of acquiring ADP resources keep pace with changes in technology. This concludes my statement, Mr. Chairman. I will be happy to respond to any questions you or other members of the Subcommittee may have. 10 A'PTACHMEZW! I A!TTACHMl%fT I Total Procurements ($1943.1M) I- Other ($355.3M) 82%- - Compatible ($1,587.8M) ,,.-l -1 _,- 1.’ 11 A?TAoIMENTI A!JTAC7!MENT I Total Procurements According to vpe of Compatlblllty ($1943.1 M) 9 Unisys-Compatible ($387.3M) 1 [- zfer Compatible ($10.1 M) Non-Compatible ($3553M) 3% Control Data Corporation- Compatible ($53.9M) 6% Honeywell Bull- Compatible ($123M) I IBM-Compatible ($1013.5M) A’ITACHMEt?T I A’JT’A(lXMENT I Total Procurements According to Manufacturer ($1943.1 M) 5% National Advanced Systems ($102.&l) 3% Storage Technology Corporation ($54.7M) Unisys ($382.6M) 4% Other ($76M) - 8% Amdahl($157.6M) 5% Control Data Corporation ($96.9M) 8% Honeywell Bull ($147.6M) IBM ($922.9M) . Memorex (representing 1 percent or $Q.6M) and NCR Comten (representing 1 percent or $17.6M) are included in the ‘other’ category. Adjustment Made Due to Rounding 13 ATFA- 1 A’I’I’AcIHMfDF I Total Procurements According to Procurement Method ($1943.1 M) 7 %I Schedule Purchases ($77.1 M) Other ($32.5M) New Contract-Sole Source ($186.1 M) 5% New Contract-One Offeror ($109.1 M) New Contract-More Than One Offeror ($243.1 M) New Contract-Developer or Integrator ($418.6M) 1% New Contract-E(a) Firm ($1&4M) Modifications to Existing Contracts ($85&2M) Adjustments made due to rounding. 14 AlTAm 1 A!lTA(ZHMENTI Compatlblo Proawmonts According to Type of Compatlblllty (81587.8M) Unisys-Compatible ($387.3M) Other Compatible ($10.1 M) sroo)ata Corporation- Compatible . 0% Honeywell Bull-Compatible ($123M) 64% - - IBM-Compatible ($1 ,013.5M) ,’ 15 ATTACHMENT I ATTACHMENT I ‘- Compatlble Procurements Accordlng to I I ,.) . ..*I- ‘I %A Procurement Method ($1587.8M) 5% GSA Schedule Purchases ($72.4M) 2% Other ($31 ,l M) New Contract-Sole Source ($165.4M) 7% New Contract-One Offeror ($103.9M) New Contract-More Than One Offeror ($213.8M) New Contract- Developer or Integrator ($203.8M) [ * L i% Contract-t)(a) Firm ($17.4M) Modifications to Existing Contracts ($780M) J 16 . A’l’T- I IBM-Compatible Procurements According to Manufacturer ($1013SM) Memorex ($9.5M) National Advanced Systems ($102.5M) 1% NCR Comten ($12.7M) Storage Technology Corporation ($54M) Amdahl($l51.9M) . 65% l A - IBM ($662SM) Adjustment Made Due to Rounding 17 L IB M - C o m p a tlble Procurements According to P r o c u r e m e n t M e thod ($1013.5M) 6% G S A S c h e d u l e P u r c h a s e s ($62.7M) 1% O ther ($10.3M) 7 2 Contract-SoleS o u r c e ( $ 3 7 M ) -j’ 7% ‘o r N e w Contract-OneO fferor ($69.7M) 1 6% ' N e w Contract-MoreT h a n O n e O fferor w ($163.8M) f : 19% -.. - N e w Contract- D e v e l o p e ror Integrator ($193.3M) 1 \‘. L $z Contract-B(a) Firm ( $ 1 7 M ) Modifications to Existing Contracts ($459.7M) 18 t ATTACHMENT I ATTACHMENT I , ---_ .. Total Procurements According lo Agsnoy and Marahcturer (S1943.1M) mo @J--h- . tn 19 l ATTACHMENT I ATTACHMENT I Agrlcutture Procuremenls Adbrdhg to HIS Procurements Aaxwdhg to )Ibnufecturer (S64.6M) Melufecturer (S193.5M) 23% Unisys ($1 su) I 6.6% Honeywell Bull #4.4M) tBM (SlSl .7M) IBM (S53.2t.4) NASA Procuremenls Manufacturer ($1044M) According lo -- I -.- 6.9% Unisys (S7.3M) Treasury Procurements Manulaclurer (S270.9M) According to 9.7% Unisys (S26.2M) Other (S15.M) Other w2.gM) 2% Amdahl (S13.3M) XaM(s5J-w Honeywell Bull ($1 M) Honeywell But! (S12.7M) IBM ($195.2M) ATTACHMENTII ATTACHMENT1.1 RELATEDGAO PRODUCTS ADP Preent. Contractgm0 and M-et Share uormatlon . (GAO,IMTEC-90087FS;Aug. 30, 1990). ce ADP Proclyllletgent. Contractma. and Market Share Informat ion (GAO/IMTEC-90077FS, July.31, 1990). cultyre ADP Procurement Contractma. and Market Share Infofinat;ion (GAO/IMTEC-90-CiFS, June 27, 1990). e Information NASA ADP Procurement. C ntractins and Market Share Information (GAO/IMTEC-90-39FS, & .O20, 1990). . Force ADP Procurement . Con ractmca and Market Shar e Infonmation (GA0,1MTEC-90i35FS,tA&m. 9, 1990). Procurew nt. . (GAO/IMTEC-90-28FS, ,ar."l,rlz9ip. Nave ADP Procurgg@mt. C tractina and Market Share Information (GAO/IMTEC-89066-FS,'Sep~~ 15, 1989).
Competitiveness of Federal Computer Procurements
Published by the Government Accountability Office on 1990-09-13.
Below is a raw (and likely hideous) rendition of the original report. (PDF)