Further Development of Navy’s Source Data System Needs to IE3e Reassessed w I 141355 (,AO,/IM’L‘E:(:-t)O-‘Lr, United States General Accounting Office Washington, D.C. 20548 Information Management and Technology Division B-238255 May 81990 The Honorable John P. Murtha Chairman, Subcommittee on Defense Committee on Appropriations House of Representatives Dear Mr. Chairman: This report responds to your predecessor’sSeptember 16, 1988, request, and discussionswith your office, to review the Navy’s development and management of the SourceData System project (SD@. Twelve years ago the Navy initiated this project, with life-cycle costs initially estimated at $1 billion, to automate military pay and personnel data reported by its field activities. The SDSproject automates a manual processof preparing and editing data entry documents and transmitting the data to the Navy’s centralized systems for pay and personnel processing.These central systems are used to pay about $21 billion annually to nearly 704,000 active duty and drilling reserve personnel. The primary objective of the SDSproject is to increase the accuracy and timeliness of the data reported to these central systems. SDSwas to be developed and deployed as three separate subsystems:one for active duty personnel in the United States and overseas;a secondfor active duty personnel on-board ships; and a third for reserves.’ These three subsystemsand the telecommunications connecting them to the Navy’s centralized pay and personnel systems were to provide a Navy- wide solution to automating the manual input of pay and personnel data reported by field activities. Our objectives, scope,and methodology are detailed in appendix I. Results in Brief The Navy’s SDSproject has fallen far short of original expectations. As of the end of fiscal year 1989, the Navy had spent about $174 million and still had not completed the project. While SDSwas to have been com- pleted in 1982, only about 64 percent of active duty personnel had been converted to SDSas of February 28,199O. Active duty personnel sta- tioned overseasand on-board ships are not on SDS. The Navy estimates v that SDSwill be deployed Navy-wide for active duty personnel by 1994, ‘The reserve subsystem was to support only the drilling reserves. Drilling reserves are defined as ready reserve personnel who participate in a Na3al Reserve program in a drill pay status. Page 1 GAO/IMTECBO-25 Navy’s Source Data System B238265 over 12 years later than originally planned. The Navy now estimates that it will cost an additional $33 million to complete development and deploy SDS Navy-wide. The reserve subsystem is no longer part of the SDSproject since its requirements are being incorporated into a sepa- rately developed system. At those locations where snsis deployed, reporting of personnel and pay data is considerably more accurate and timely. However, SDS does not meet all performance goals where deployed, nor will it achieve all the expected performance goals when fully deployed Navy-wide. The program has been characterized by unclear lines of authority and responsibility for developing and deploying SDS, failure to coordinate among Navy commands,delays, and false starts. While original SIX plans called for a senior-level steering committee to coordinate the development and deployment of SDSacrosscommands,this committee never convened.The authority and responsibility to develop and deploy SDSis still fragmented within the Navy, and there is no current Navy- approved plan, strategy, or economic assessmentof the overall SDS pro- ject. Further, the oversight provided by the Office of the Secretary of Defensewas both incomplete and prematurely withdrawn. Becauseof this project’s long history, fragmented authority, and lack of oversight, we were not always able to obtain complete information or identify the precise causesfor the problems we found. Nevertheless,it is clear that this project has wandered and needsto be reevaluated. We are making recommendations aimed at reassessingthe overall management and future direction of this project and curtailing further investments in it while a reassessmentis made. Background The Navy’s personnel and pay operations are split between two major commands.The Naval Military Personnel Command, operating under the direction of the Chief of Naval Operations, is responsible for person- nel management functions. The Navy Accounting and Finance Center, operating under the direction of the Comptroller of the Navy, is respon- sible for military pay functions. To carry out their respective personnel and pay missions, these Navy activities must share information in their centralized automated systems becausemany personnel actions estab- lish, change,or discontinue certain military pay and allowances. In the late 1970s the Navy primarily used mail-based, manual reporting systems for transmitting pay and personnel data from field activities to centrally located, but separate pay and personnel systems. Between Page 2 GAO/IMTEG9O-26 Navy’s Source Data System . E228255 1978 and 1986, we, the Naval Audit Service, and the Naval Inspector General reviewed the Navy’s field reporting and centralized pay and personnel systems.2The common theme of these reports was that the exchangeof data between field disbursing offices and the central pay system and between field personnel offices and the central personnel system was inaccurate and slow. In fact, the Navy found the central pay system to be inaccurate about 48 percent of the time in 1979. As a result, the Navy, to pay the service members accurately on payday, paid different amounts than the amounts shown on the central pay system. In 1982kwereported that the Navy’s pay system did not meet generally acceptedaccounting standards and recommendedthat the Navy bring its system into compliance with these standards.3The Navy, as required by the Federal Managers’ Financial Integrity Act,* reported to the Presi- dent and the Congressin 1984 that its military pay accounting system did not conform with the principles, standards, and related require- ments prescribed by the Comptroller General. In fiscal year 1989, the Navy reported that corrective actions had been taken and the pay sys- tem substantially conformed to these requirements. To improve its military pay and personnel systems, the Navy began, in 1977, a major program called the Pay and Personnel Administrative Support System (PASS) to streamline the data reporting processand improve communications between the personnel and disbursing field offices and the centrally located pay and personnel systems. This two- phased program was co-sponsoredby the Comptroller of the Navy and the Chief of Naval Operations and was to be managed and implemented under the direction of the PASS program manager within the Personnel Command. The first phase called for merging field pay and personnel organizations. By 1981, approximately 3,600 personnel offices and 134 disbursing offices ashore were consolidated into 162 pay/personnel support offices worldwide. The secondphase, SDS,called for developing an automated field report- ing and management information system to support the consolidated field offices as well as the ship-board offices. The field reporting system 2Seepage 17 for a list of related GAO and Navy reports. 3GA0 letter to the Secretary of the Navy, B-199833, Feb. 2,1982. *Federal Managers’ Financial Integrity Act of 1982,314J.S.C. 3612 (b) and(c). Page 3 GAO/IMTECBO-26 Navy’s Source Data SYStem B-288266 was to be developed and deployed as three separate subsystems,each supporting a different personnel community-active duty personnel stationed in the United States and overseas(referred to as ashore), per- sonnel on-board ships (referred to as afloat), and personnel in the drill- ing reserves.The managementinformation system was to provide field offices accessto data in the central pay and personnel systems that were not available under the manual reporting system. The Navy decided to develop the ashore subsystem first, then the reserve subsystem, and finally the afloat subsystem. According to SDS project office officials these three subsystemswere to be developed sep- arately to accommodatethe different hardware and communications environments employed by each user community. The SDS project was to be a joint effort of the Personnel Command and the Comptroller of the Navy. An SDSproject managementoffice was established within the Personnel Command to design, develop, and deploy sns. SDSProject Hindered develop, deploy, and operate sns,but this effort has fallen far short of by Poor Planning and original expectations. As of February 28, 1990, only about 64 percent of Management active duty personnel had been converted to SDS. Active duty personnel stationed overseasand on-board ships are not on SDS,and reserves are no longer part of the snsproject. Development of The Navy spent over $13 million developing reserve and afloat subsys- SubsystemsContinually terns but they were not used. Also, deployment overseasof the ashore subsystem was delayed becauseof a lack of funds. Changing Reserves.After spending about $1.1 million to nearly complete develop- ment of a reserve subsystem, the Navy decided to satisfy the reserve’s requirements with another system being developed by the Commander, Naval ReserveForce. Consequently, work on the subsystem was suspended. Afloat. The Navy prototyped the afloat subsystem at a cost of $12 mil- lion on 12 ships for about 6 years. In September 1989, however, the Navy abandonedthis effort. Page 4 GAO/IMTEGBO-25 Navy’s Source Data System 4, E228266 The SDS project office has recently prepared conceptual plans for devel- oping and deploying a new afloat subsystem, which is expected to cost at least $24 million. This subsystem will be developed in two modules. The first module will utilize an already partially deployed scaled-down interim system called the Uniform Microcomputer Disbursing System (UMIDS)~ for reporting pay data. A secondmodule will be developed for reporting personnel data. Over the long term the Navy plans to evolve these two modules into a single integrated afloat subsystem, which they plan to interface and/or operate with a new ship-board hardware plat- form planned for the mid-1990s. Life-cycle costs for the new approach are not yet known. The Navy expects to make a deployment decision on this approach in November 1990. OverseasDeployment of Ashore Subsystem.Deployment of SDS for active duty personnel stationed overseasalso encountered delays and redirection. An ashore subsystem was developed to support both United States and overseaslocations. According to the project’s original plan, deployment of SDS overseaswas to have been completed in 1982. How- ever, according to the Navy, the overseasdeployment was delayed becausesufficient funding was not available. In the meantime, according to the project manager, the manufacturer has comeout with a new, less expensive model of the hardware previously planned for deployment overseas.This hardware is being used along with the already developed ashore software to prototype the subsystem for overseasactivities. This approach is expected to cost at least $9 million. Subject to the availabil- ity of funding, the SDS project manager expects to commencedeployment overseasin December 1991. Life-Cycle Costs and Since the inception of SDS, the Navy has had difficulty identifying life- Benefits Have Fluctuated cycle costs and benefits. From 1977 to 1985, the Navy prepared four separate economic analyses of life-cycle cost and benefit estimates justi- Widely fying the SDS project6 The life-cycle cost estimates changed significantly, starting with costs estimated initially at $1.1 billion in 1977, then declin- ing to $869 million 2 years later, then increasing to over $3 billion in another 2 years, and most recently to $643 million 4 years later in 1986. At the sametime, the Navy’s cost estimate for developing and deploying ‘UMIDS was developed during 1981 and 1982 as an interim system until SDS was deployed. UMIDS was designed to support only pay functions, not personnel. “These economic analyses did not include estimates of costs and benefits for the afloat subsystem. Also, the 1979 economic analysis did not include estimates of costs and benefits for the reserve subsystem. Page 6 GAO/IMTlKXO-26 Navy’s Source Data System the ashore and reserve subsystemssteadily increased from $13.8 million in 1977 to over $180 million in 1986. Estimated benefits, on the other hand, remained relatively unchanged in the first three economic analy- ses,but increased five-fold in the latest estimate, made in 1985. We could find no evidence that the Navy had independently verified these analyses. Despite these wide swings, Defenseand Navy officials allowed the project to proceed. Furthermore, even though the SDS project has undergone a number of delays and changesin direction since 1985, the Navy has not updated its estimates of costs and benefits. Project Management and No single organization has been responsible for managing SDSNavy- Control Was Not Adequate wide. Initially, the three subsystems(ashore, afloat, and reserves) were to be developed under a single project office coordinated by a senior steering committee. However, at least three different offices have been responsible for developing them, a fourth office developed the UMIDS system, and no office was clearly responsible for ensuring a coordinated approach. In fact, we had considerable difficulty getting information such as cost data for all three subsystemsand the UMIDS system because there was no single point of contact that was able to identify life-cycle costs or actual expenditures for SDSNavy-wide. The program has been characterized by unclear lines of authority and responsibility for developing and deploying SDS,failure to coordinate between commands, and delays. Although initial planning documents called for a senior-level steering committee to coordinate the develop- ment and deployment of SDSacrosscommands,we found no evidence this committee ever convened.As a result, divided and ever-changing organizational and project development responsibilities did not ensure a focused, timely solution to the development and deployment of SDS Navy-wide. The Navy recognizedthese problems in 1989. As a result of a program review of the afloat subsystem, the Navy concluded that the program managementwas diffused, only part of the program was being reviewed by various project managers,and coordination among the various orga- nizations was lacking. The Navy also concluded that management defi- ciencies were created, in part, by a June 1983 decision that relieved the PASprogram manager of any responsibility for the SDSproject. The Navy review made a number of recommendations that included: . redefining the authority, role, and responsibility of the PASSprogram manager as intended in the original functional description; Page 6 GAO/IMTEC9O-26 Navy’s Source Data System E-238256 . establishing a steering group or committee to provide coordination with various commands; l developing a memorandum of understanding to define clear lines of authority and responsibility; and l realigning headquarters elements and resourcesto support an effective program managementorganization. As of March 9,1990, the Navy had developed a memorandum of under- standing to return responsibility for developing the afloat subsystem to the SDS project office. Also, a project managementcharter is being drafted to define lines of authority and responsibility for developing and deploying the afloat subsystem. The Office of the Secretary of Defense’sMajor Automated Information SDSProject Lacked System Review Committee (MAISRC)~ had oversight responsibility for SDS Sustained Defense from 1979 to 1987. During its initial review in 1982, the MAISRC stated Program Oversight that the Navy’s proposed development approach for the ashore subsys- tem was sound and, in developing SDS, it should continue to apply sound life-cycle managementprinciples and ensure the availability of adequate telecommunications and acquire and upgrade, as necessary,hardware and software resources.The MALSRC also directed the Navy to furnish it with any future plans for developing the reserve subsystem and to notify it of slippages in the project and ensure that privacy require- ments and security safeguards were incorporated. However, MAISRC records did not indicate that there was any concern over the estimated cost or benefits of the project. In 1987 the MAISRC, satisfied with the development and deployment schedule of SDS for ashore and reserves,delegated its oversight responsi- bility of SDS to the Navy. The MAISRC, however, did not review the afloat subsystem as part of the overall SDSproject, even though it had always been a major part of SDS. According to the SDS program analyst from the DefenseComptroller’s Office, the MAISRC did not review the afloat sub- system becauseit was being developed and funded as a separate pro- gram under the Ship-board Non-Tactical ADP Program (SNAP). After the MAISRC delegated oversight to the Navy, the Navy redirected the afloat prototype for personnel assignedon-board ships and the ‘This committee was created within the Office of the Secretary of Defense to provide structured oversight and prudent fiscal management in acquiring major automated information systems. Page7 GAO/I~~~TEC~O-~~N~V~'~SOU~C~DataSYfhn . El-238255 planned utilization of the ashore portion for personnel stationed over- seas.As noted earlier, the Navy is making a number of changesto its conceptual approach for providing SDS to personnel assignedon-board ships and overseas. In mid-1988 the Assistant Secretary of Defense(Comptroller) required each military service to report quarterly on all systems for which it had delegated oversight review. Although the project office prepared these reports, the Navy did not file the first four of the required quarterly reports since the ashore subsystem had already been developed and was being deployed. The Comptroller’s office initially agreed,but later required the Navy to begin submitting reports as a result of our ongoing audit work. The Navy submitted its first quarterly report on SDS for the period ending September30,1989. At those locations where SDS is deployed, reporting of personnel and pay SDSImproves data is considerably more accurate and timely. However, SDS does not Accuracy and meet all performance goals where deployed nor will it achieve all the Timeliness but Some expected performance goals when deployed at the remainder of the intended locations. Further, certain other performance goals established Performance Goals for SDSare unrealistic since their achievement dependson actions Are Not Achievable outside the control of SDS. The Navy established three primary goals for SDS:(1) error rates in data used to update the central pay and personnel data baseswere to be less than 3 and 1 percent respectively, (2) 99 percent of the field-generated pay and personnel transactions were to be transmitted and prepared for updating within 24 hours of releaseby the field offices, and (3) no more than 1 percent of all centrally computed pay amounts would be overrid- den at field offices within the continental United States. Prior to SDS the update error rates were 9 and 13 percent for pay and personnel respec- tively; on average,it took 13 days to send data, via mail, from the field offices to a central office and to prepare the data for updating the cen- tral pay and personnel systems; and the override rate was 48 percent. At those field activities where SDS is deployed, Navy statistics show that update error rates and untimely submissionsdecreasedsignificantly from May through November 1989. SDS error rates of 1.3 and .Opercent for pay and personnel data respectively were well within the estab- lished goals. For timeliness, about 89 percent of the pay and 86 percent of the personnel inputs were transmitted and ready for updating within 24 hours, which is short of the 99 percent goal. The Navy also reduced Page 8 GAO/IMTEG9O-25 Navy’s !3ource Data System its override rate from 48 percent to about 10 percent, which is short of its 1 percent goal. Another goal was to reduce the annual loss resulting from overpaying personnel when they leave or separate from the Navy to less than $4 million. Navy records showed that although the number of personnel being overpaid has been reduced, these lossesrose from $6 million in 1979 to an estimated $16 million in 1989. The $4 million goal, however, cannot be achieved unless changesunrelated to SDSare made. One major source of overpayments is the payment of reenlistment bonusesto per- sonnel who do not complete their commitment. While SDScan assist in the identification of funds owed by service members who do not com- plete their enlistment, the collection of any funds due the government, which often requires a manual collection process,is outside the control of SDs. Clearly the Navy needsto fix its systems for field reporting of pay and Conclusions and personnel data-we cameto this conclusion long ago and so did the Recommendations Navy and its auditor and inspector general staffs. When properly planned and applied, modern technology offers viable solutions to make these fixes. Yet, after spending 12 years and about $174 million, the Navy haa produced only a portion of the originally envisioned system. Where the system has been deployed, it has produced substantial improvements in the accuracy and timeliness of reported data; however, the rest of the Navy is not reaping these improvements. Furthermore, the Navy has not prepared realistic estimates of life-cycle costs and ben- efits in the past nor have such estimates been prepared for its new strat- egy for providing SDS support for personnel on ships and overseas. Technology still offers potential benefits to the Navy, but until the causesof SDSdelays, cost escalations, and false starts are identified and corrected, the public, the Congress,and the Secretary of Defensecannot be assured that further expenditures of time and funds on SDS will be wisely spent. Project oversight by the Office of the Secretary of Defense and the Navy is necessaryto ensure that these problems are corrected and the SDS program produces cost-effective solutions in a timely manner. We recommend the Secretary of Defensedirect the Secretary of the Navy to (1) reassessthe management,plans, goals, alternative solutions, Page B GAO/IlWEG90-25 Navy’s Source Data System B238288 cost, and benefits of the SDSproject and determine whether it or a por- tion of it should be continued, and (2) ensure that spending for SDSand planned solutions are held to a minimum while this reassessmentis conducted. If, on the basis of this reassessment,the Navy decidesto continue with SDS, we recommend that the Office of the Secretary of Defensereview the Navy’s decision and reinstate MAISRC oversight for the entire project. In addition, we recommendthat the Office of the Secretary of Defense provide the Congresswith a revised budget forecast for SDSbased upon a current estimate of project and life-cycle costs. As requested by your office, we did not obtain official agency comments on this report. We discussedthe issuesin this report with officials from the Department of Defenseand the Department of the Navy, and have included their comments where appropriate. As arranged with your office, unless you publicly announcethe contents of this report earlier, we plan no further distribution until 30 days from its issue date. We will then send copies of this report to the Chairmen, House and SenateCommittees on Appropriations, and other interested Members of Congress;the Secretariesof Defenseand the Navy; the Director, Office of Management and Budget; and others upon request. This work was performed under the direction of Samuel W. Bowlin, Director, Defenseand Security Information Systems,who can be reached at (202)-275-4649.Other major contributors are listed in appen- dix II. Sincerely yours, t!syyJWL w Ralph V, Carlone Assistant Comptroller General Page 10 GAO/IMTEG90-26 Navy’s Source Data System Page 11 GAO/IMTEGBO-26 Navy’s Source Data System Contents Letter Appendix I Objectives, Scopeand Methodology Appendix II 16 Major Contributors to This Report Related GAO and 20 Navy Products Abbreviations GAO General Accounting Office IMTEC Information Management and Technology Division MAISRC Major Automated Information System Review Committee Y PASS Pay and Personnel Administrative Support System SDS Source Data System SNAP Ship-board Non-Tactical ADP Program UMIDS Uniform Microcomputer Disbursing System Page 12 GAO/IMTECXO-25 Navy’s Source Data System Y Page 13 GAO/IMTECXO-25 Navy’s Source Data System Appendix I Objectives,Scopeand Methodology On September 16,1988, the Subcommittee on Defense,House Committee on Appropriations, requested that GAO review the Navy SourceData System (SDS). In subsequentdiscussionswith the subcommittee staff, our specific objectives were to determine whether the project is meeting expected objectives and goals and to identify any shortfalls of the pro- ject. We agreed to determine the original and present scope,goals, and objectives of SDS and identify reasonsfor any changes;to evaluate and measure the extent to which SDS is meeting stated goals and objectives; and to identify other automation initiatives being pursued by the Navy that may addressSDS goals and objectives. We examined the original project scopeas presented in the 1977 Eco- nomic Analysis and compared it with other life-cycle management docu- ments. In addition, we met with cognizant agency officials in order to determine the reasonsfor changesin the scopeof the project and to identify other automation initiatives started by the Navy and to also have them clarify the performance goals and objectives. We obtained various performance statistics from the Navy Finance Center and from the Naval Military Personnel Command and compared those statistics with the stated goals. In addition, we reviewed the pay and personnel operations at 10 PersonnelSupport Detachments in order to determine how SDS operates and determine if it is meeting its perform- ance objectives. We also met with agency officials to obtain their com- ments on our findings. Our review was conducted primarily at the Navy Accounting and Finance Center and the Naval Military Personnel Command in Washing- ton DC. In addition, we interviewed agency officials at the Navy Finance Center in Cleveland, Ohio. We also interviewed personnel aboard the USSO’Brien and the USSHewitt at San Diego Naval Station and the USSBainbridge and USSTrenton at Norfolk Naval Station. Our review was conducted from November 1988 to March 1990. We dis- cussedissuescovered in this report with officials from the Department of Defense,Office of the Inspector General; the Office of the Comptroller of the Department of Defense;Department of Navy, Naval Military Per- sonnel Command, and the Navy Accounting and Finance Center. In doing our work, we had difficulty obtaining information on certain deci- sions or explanations for those decisions.This difficulty arose because of the long history of this project, its changing and fragmented organiza- tion structure and its lack of central control and oversight. As you requested, we did not obtain written agency comments on a draft of this Page 14 GAO/lMTEG90-26 Navy’s Source Data System Append& I Objecthe, Scope and Methodology report but did incorporate their verbal comments where appropriate. Our work was performed in accordancewith generally acceptedgovern- ment auditing standards. Y Page 16 GAO/IMTJC90-25 Navy’s Source Data System Appendix II Major Contributors to This Report James R. Watts, Associate Director Information Joseph T. McDermott, Assistant Director Management and Robert L. Cracker, Evaluator-in-Charge Michael B. Corrado, Evaluator Technology Division, Washington, DC. Donald A. Weisheit, Assignment Manager Detroit Regional Office Louis M. Ockunzzi, Site Senior Lynette A. Westfall, Evaluator Y Page 16 GAO/IMTECXM-25 Navy’s Source Data System ? Y Page 17 GAOjIMTESBO-25 Navy’s Source Data System :” .- Page 18 GAO/IMTEC-90-25 Navy’s Source Data System Y Page 19 GAO/IMTEGQO-25 Navy’s Source Data System RelatedGAO and Navy Products The Navy’s Advanced Information System - A Personnel Information System for the 1980-1990s(GAO/LCD-78-122, Sept. l&1978). The Navy’s Computerized Pay System is Unreliable and Inefficient - What Went Wrong? (GAOIFGMSD-80-71, Sept. 26, 1980). Report of the Navy Pay System Review (Naval Inspector General, 008/ 2666, Dec. 17,1982). Pay/Personnel Administration Support System (PASS) and Joint Military Entitlements Payment System (JUMPS) (Naval Audit Service, T202222, Nov. 29,1983). Survey of Actions to Correct Problems with the Navy’s Military Pay System (GAO/AFMD-86-05, Oct. 12, 1984). Navy’s ProgressIn Implementing The Federal Managers’ Financial Integ- rity Act (GAOINSIAD-86-160, Sept. 27, 1985). Debt Collection: Navy’s Efforts to Collect Debts From Former Service Members (GA~/AFMDSS-El-BR, May 19, 1986). Development of the Navy Pay and Personnel Source Data System (Naval Audit Service, D30006, Nov. 26, 1986). Y (610346) Page 20 GAO/IMTECXO-26 Navy’s Source Data System Ktqwsts for copies of GAO reports should be sent. to: I J.S. Gtw~ral Accounting Office Post Office Box 6016 Gaithersburg, Maryland 20877 ‘I’lw first five copies of each report. are free. Aclciitioual copies are !K?.oo tvwh. Thtvt~ is a 25% discount, on orders for 100 or more copies mailed t,o a single address. Orders must. be prepaid by cash or by check or mont~y order matit& out (,o t.tw Sul)t?rint.ttntient of Documents. I. S.
Computer Systems: Further Development of Navy's Source Data System Needs to Be Reassessed
Published by the Government Accountability Office on 1990-05-08.
Below is a raw (and likely hideous) rendition of the original report. (PDF)