Competing risks quantile regression at work : in-depth exploration of the role of public child support for the duration of maternity leave

Document Type

Journal article

Source Publication

Journal of Applied Statistics

Publication Date



Advance online publication




Dependent competing risks, quantile regression, quantile crossings


Despite its emergence as a frequently used method for the empirical analysis of multivariate data, quantile regression is yet to become a mainstream tool for the analysis of duration data. We present a pioneering empirical study on the grounds of a competing risks quantile regression model. We use large-scale maternity duration data with multiple competing risks derived from German linked social security records to analyse how public policies are related to the length of economic inactivity of young mothers after giving birth. Our results show that the model delivers detailed insights into the distribution of transitions out of maternity leave. It is found that cumulative incidences implied by the quantile regression model differ from those implied by a proportional hazards model. To foster the use of the model, we make an R-package (cmprskQR) available.



Print ISSN




Publisher Statement

Copyright © 2016 Taylor & Francis. Access to external full text or publisher's version may require subscription.

Full-text Version

Publisher’s Version

Recommended Citation

Dlugosz, S., Lo, S. M. S., & Wilke, R. A. (2016). Competing risks quantile regression at work: In-depth exploration of the role of public child support for the duration of maternity leave. Journal of Applied Statistics. Advance online publication. doi: 10.1080/02664763.2016.1164836