Random effects Weibull regression model for occupational lifetime

High job turnover rate can cause many problems and each company needs proper strategies to prevent the brain-drain of its manpower. For effective human resource management, predicting the occupational life expectancy or the mean residual life of those who are to leave and join another company is important. In this paper, we propose a random effects Weibull regression model for forecasting the occupational lifetime of the employees who join another company, based on their characteristics. Advantage of using such a random effects model is the ability of accommodating not only the individual characteristics of each employee but also the uncertainty that cannot be explained by individual factors. We apply the proposed model to the occupational lifetime data obtained from the company affiliated to general trading in Korea. From our analyses, we can infer the characteristics of those who have a relatively longer occupational lifetime as follows: the managing director level, relatively old, those who entered the company earlier, high school graduates, those who were involved in technical service, and married female employees. Accordingly, effective human resources management policy is necessary to retain those who are good but want to leave and those who stay but need more improvement for the betterment of the company.