Purpose: The purpose of this paper is to provide empirical evidence for individual educational investment decisions and to investigate the relative importance of two factors, the type of education (vocational vs. academic) and subject area (e.g., commercial or health), in determining variance in earnings. Design/methodology/approach: Using a sample of 1200 individuals based on the 2011 Swiss Adult Education Survey, Mincer-type earnings equations are estimated. The variance in earnings is decomposed with respect to the two factors mentioned above, which allows to quantify the relative contributions of type of education and subject area to variance in earnings. Findings:The results of the variance decomposition show that subject area explains nearly twice the variance in earnings compared with that explained by type of education. Social implications: As results show that earnings variance-and thereby risk-relate more to subject area than to type of education, this study suggests that for individuals caring about the risk of their educational decision the selection of a specific subject area is more relevant than the choice between vocational and academic tracks; in addition, educational policies as part of HRM policies should devote as much attention to the choice of subject areas as to vocational or academic education. This is especially important for companies or countries planning to introduce or to extend vocational education as part of their human resources strategies. Originality/value: This study is the first to show whether earnings vary more by type of education or by subject area.