This paper provides an empirical description of self-employment duration in the presence of two competing risks: wage-employment and non-employment. Using high-quality Dutch administrative panel data from 2000 to 2015, we show several risk factors for survival rates. Special attention is paid to the roles of gender, cohorts and industries. The main findings are (1) men survive longer than women, because men are less likely to become non-employed, (2) becoming a father implies a lower risk of non-
employment, while becoming a mother implies a higher risk, (3) having a partner is related to survival among men, but not among women, (4) survival rates increase substantially with each successive cohort, (5) compositional eects in cohort dierences are unlikely, (6) self-employment serves as a stepping stone to wage-employment in consultancy and IT, but not in other industries. These results are fairly robust to parametric models and alternative denitions of self-employment.
Keywords: Administrative data; Competing risks; Entrepreneurship; Self-employment; Survival analysis