Forecasting Enlistment Supply: A Time Series of Cross Sections Model

Because the military relies on voluntary enlistments of fill its entry-level positions, there has been widespread interest in estimating how military enlistments respond to various supply factors, both those originating in the civilian sector (e.g., business cycles) and those over which the military exerts some control (e.g. recruiters). To help the military anticipate manpower shortages before they actually develop, these estimated supply parameters are used to predict the future course of enlistments under various hypothetical situations. This report documents research on a model of the supply of high aptitude, high school diploma graduate, non-prior service male enlistees. The emphasis is on methodology, including variable construction and methods of estimation and forecasting. The methodology is applied to monthly state-level data over the period October 1974 through March 1981. The output is fitted equations for the four Services that relate the enlistment rate to military/civilian pay, the number of recruiters per potential enlistee, a business cycle variable, and other control variables reflecting changes in enlistment policy, including the end of the GI Bill. The fitted models are then used in conjunction with future scenarios to obtain forecasts of high quality (AFQT category I-IIIA, high school graduate), non-prior service male enlistments.