Regression Analysis of Demand for U.S. Military Labor

The past year has been one of unprecedented change for the U.S. military organization. It would be of interest to analyze the impact of the changing situation on U.S. military labor demand. In this thesis, several demand models for U.S. military labor are considered to identify influential factors that predict the size of future military labor. A stepwise regression analysis is used to select some significant demand models. Data used to construct demand models in this thesis over the period of 1963-1986 while actual data (1987-1990) are used to check model validity. Demand models selected are used to stimulate force levels through the year 2000 under various scenarios. Major contributions of this thesis are employing international security (war casualties), USSR's military-end-strength and the past year's U.S. end strength in the prediction models for military labor demand. The resulting demand models turn out to be more parsimonious but they have better predictions that existing models. The results of simulation study based on various scenarios regarding war casualty and USSR's end strength would provide policy makers with useful information for future defense manpower plan.