Data envelopment analysis models and structures and their use in the management of Army recruitment

This dissertation develops a new Forecasting and Allocation of Army Recruiting Resources Study-Sequential Hierarchical Allocation of Resource Elements (FAARRS-SHARE) methodology for forecasting at the U.S. Army Headquarters (HQDA) level least cost allocations of resource elements to assist in downsizing policy determinations. This new methodology, for the first time, represents actual recruiting operations tied into the budgeting process. The new semi-parametric multiplicative Data Envelopment Analysis (DEA) stat-based methodology provided in this dissertation allows for synthesis of the relatively high resolution battalion level data together with accounting data in order to form aggregate policy determinations. The implementation of this methodology within the U.S. Army hierarchy also opens and coordinates communications between several different operating levels of the Army personnel management organization. This is possible since PC implementation of the methodology is simultaneously available and operated at the HQDA, U.S. Army Recruiting Command (USAREC) and other relevant offices.; These new models and structures require accurate software to identify the efficient facets of the Army recruitment production function from empirical data. Accordingly, the process of identifying the members of the efficient facets for the additive model of DEA is investigated, and some procedures for enhancing this identification are proposed and tested.