Summary

The Defense Department (DOD) has been reducing the costs of civilian personnel management during the past five years by reducing the number of staff working in personnel, consolidating personnel management functions at newly created regional centers, and trying to improve personnel management business processes. A key part of this effort is DOD's development of a new information management system--the Defense Civilian Personnel Data System--to support a wide range of personnel management duties, including recruitment, benefits administration, and training. DOD expects to complete deployment of this system by March 2000. This report answers the following questions about this initiative: How did DOD determine the number and locations for civilian personnel regional service centers and why is there such a wide disparity in the number of regional centers among the services? Is DOD applying the principles of the Clinger-Cohen Act in overseeing, managing, and developing the Defense Civilian Personnel Data System? Does that system replicate a system that is available through the Office of personnel Management? Was DOD leadership aware of the extent and cost of the needed modifications to the commercial off-the-shelf software application? Has DOD identified and mitigated the risks associated with the major commercial off-the-shelf modifications?

GAO noted that: (1) DOD's current initiative can potentially improve civilian personnel operations and achieve cost savings; (2) however, because the Department has not examined other business process alternatives that could have potentially achieved even greater savings and process efficiencies, there is no assurance that this is the best alternative for civilian personnel operations; (3) before embarking on its costly initiative to improve personnel management, DOD examined two alternatives: (a) outsourcing personnel computer operations to the Department of Agriculture's Finance Center; and (b) regionalizing personnel centers; (4) DOD determined that it would take the National Finance Center about 6 years to prepare for transferring computer operations and that some new functionality built into its legacy system would be lost; (5) however, DOD did not examine several other potentially effective alternatives, including: (a) continuing to centralize all or parts of its personnel management operations to reduce duplicative layers of oversight at the components and ensure more consistent operations DOD-wide; (b) integrating its personnel and payroll management systems; (c) restructuring its regional offices to serve multiple components rather than perpetuating regional offices dedicated to only one component; (d) restructuring local personnel offices to serve multiple bases or installations (they now serve only one base or installation); and (e) outsourcing all civilian personnel operations to the private sector; (6) these alternatives are feasible and may have helped DOD to achieve even greater savings and efficiencies than the current approach; (7) in addition, the Defense Science Board determined that integrating payroll and personnel systems was a viable and cost beneficial option for military personnel; (8) the Civilian Personnel Management Service (CPMS) officials who were responsible for the personnel initiative said that they did not consider these business processing alternatives because: (a) CPMS did not have authority to require the military services and DOD agencies to adopt such approaches; (b) DOD did not allow sufficient time to rigorously examine alternatives; and (c) DOD lacked basic cost and performance data needed to study the alternatives; (9) after it decided on its approach, DOD did not follow a sound process for selecting regions; (10) DOD did not adequately consider a full range of technical options before deciding to replace its legacy system with the Oracle COTS product; and (11) after DOD acquired the Oracle system, it did not mitigate critical technical risks.