| The purpose of this research was to gain a better understanding of a government-wide, multi-agency enterprise of over thirty Federal agencies (the Agencies) engaged in declassification of government documents. Interactions among the Agencies are driven by the need for each agency with a vested interest in a document to have an opportunity to review that document prior to its release by the holding agency. These interactions are referred to as "equity reviews". The study began with development of a framework for the interactions among the four largest of the Agencies. The structure of the declassification operations within each of these four agencies was described in terms of the following effects or entities: workflow, personnel, communications, and technology. The objective of the research was to identify the principal parameters that directly affect the Critical Success Factors (CSF's): (1) declassification productivity for each agency and government-wide, (2) individual agency overall costs, and (3) process security. A base model was constructed to present the enterprise as it currently exists. Then, a dynamic analysis was carried out using the base model and three alternative configurations to represent different ways that the Agencies could interact to accomplish their shared responsibilities. The base model and its three alternatives also represent the current declassification process and a feasible migration path from fully decentralized to centralized declassification operations, as described by the Congressional Moynihan Commission. These four simulation models were then used for dynamic configuration and sensitivity studies over the expected ranges of structural and policy changes for the interactions among the agencies. Important aspects of the study were the inclusion of (1) special handling for duplicate documents detected at the log-in phase of operations, and (2) use of Just-in-Time (JIT), Inventory Control Theory, and Queuing Theory as process balancing strategies. The dynamic analysis also included statistical models to represent process and personnel characteristics, including document size and arrival frequencies; human experience levels, reading speeds and error rates; and duplicate document arrival frequencies. Results are presented as comparisons of the CSF's for the four models, over the expected operating conditions. Finally, the study provides recommendations on how to meet the government declassification requirements in the most cost-effective manner. |
