SOEPpapers on Multidisciplinary Panel Data Research 897
In this paper, we run regression analyses to explain voluntary turnover intentions with data from more than 5,000 employees and with about 250 explanatory variables. The findings of our multi-factor approach highlight the fact that previous empirical research might have over-estimated the impact and significance of many factors. We show the relevance of the so-called omitted variable bias to our findings and present an empirical approach to gain estimations that are more accurate. Our approach can estimate in detail the relevance of different factors from the following categories: (a) employees' satisfaction, (b) industry, (c) firm size, (d) status and position, (e) commuting and working hours, (f) income, incentivesand fairness, (g) career development, (h) health, (j) political orientation, (k) demographics, and (l) personality traits.