Factors affecting employees’ duration in jobs in Ho Chi Minh City, Vietnam

Current employee turnover surveys focus on predicting turnover by measuring demographics, work environment, organizational commitment, and alternative job opportunities. The analytical technique commonly used in these studies is logistic regression that does not incorporate time as a variable of interest. Therefore, this paper uses a survival analysis technique with a time variable to test the turnover model with a sample of 312 employees from various industries in Ho Chi Minh City. The estimates of the survival function and the risk indicate that with the increase in time, the likelihood of staying with the company decreases sharply in the short term, but after a particular time, the risk of leaving the job disappears. The analytical method is selected through hypothesis evaluation of Cox PH model, distribution model test. The test shows that the parametric method with exponential distribution provides a rationale analysis for the turnover model. In particular, factors such as the number of dependents, years of experience, management interest, salary, and emerging job replacement opportunities significantly predict turnover while happiness at work does not. The results from the parametric model regression are consistent with previous studies on work environment factors, organizational commitment, and alternative job opportunities that have a strong influence on employee turnover. The implication of these findings is to help organizations plan and implement employee retention policies that are appropriate over time.