A Microdata Model of Delayed Entry Program (DEP) Behavior

The Manpower and Personnel Policy Research Group of the U.S. Army Research Institute examines personnel issues of particular importance to the Army. Personnel losses from the Delayed Entry Program (DEP) is one such issue. In this paper a model is developed to predict DEP loss. The model will provide an increased understanding of the DEP loss problem along with valuable information concerning identification of individuals most likely to become losses. Two DEP loss models are created: one including high school graduates and nongraduates and a separate model for high school seniors. Maximum likelihood logistic regression (logit) estimates are make from individual data for the first half of FY82 and FY83. Both individual characteristics and policy variables are used in the analysis. These include age, gender, race, AFQT score, education, contracted DEP length, training MOS, region of the country, and enlistment and incentive options. Scenarios are staged to measure the effect of different combinations of relevant variables.