An empirical analysis of enlistment intentions and subsequent enlistment behavior

Abstract

This research uses data from the Youth Attitude Tracking Survey and the Defense Manpower Data Center to predict interest in joining the military service (propensity) for the prime market of 17 to 21 year old high school diploma graduates that are expected to score above the fiftieth percentile on the military entrance examination. A follow-on analysis of actual conversion of propensity to enlistment action is also conducted. In predicting military interest, the independent variables were restricted to those that have data available on a regional level. This will enable military recruiting commands to develop regional estimates of propensity. Multinomial logistic regression was used to estimate the interest prediction equations for population groups by race and gender. Interest categorization was possible with reasonable accuracy using local unemployment level, parent's education and the regional 'go to college' rate as the independent variables. Conversion of military interest to enlistment action does appear to vary by interest level. Follow-on research and recommendations are provided.