“Timely” Methods: Examining Turnover Rates in the U.S. Military

Abstract

Despite a long history of research on organizational turnover, questions still linger about relative contributions of different factors to the probability and timing of turnover. Complicating matters, civilian and military differences can moderate important relationships among the factors. In this study, several event history models for predicting voluntary turnover in the U.S. military were estimated. Turnover predictors included background variables, military satisfaction, organizational commitment, withdrawal intentions, job withdrawal, and comparisons of military and civilian work and lifestyles. Results showed that withdrawal intentions, job withdrawal, organizational commitment, and military tenure consistently predicted voluntary turnover. Inconsistent findings for job satisfaction and comparisons of military and civilian work and lifestyles raise questions as to their roles as predictors of military turnover.

Notes

^a M = Mean and SD = Standard Deviation (in parentheses).

^bSingle item measures; alpha could not be computed.

^cMean item imputation used for missing data.

^dComparison of opportunities in the military versus civilian world.

^aMarital status was coded as 1 (Married) and 0 (Not married).

^bGender was coded as 1 (Male) and 0 (Female).

^cRace/ethnicity was coded as 1 (Non-Hispanic, White) and 0 (Other race/ethnicity).

^dThe Service components were dummy-coded so that 1 (Given service; e.g., Army) and 0 (Not in given service).

*All significant at α = .01.