“…We added a linear and squared term for the father's age at childbirth and a categorical indicator of age of the mother at childbirth (with values: ≤25, 26/30, 31/35, 26/40, ≥41) and controlled for the birth of a second child, whether the couple was married or cohabiting, and level of education of both parents. As previous research indicates that parental leave uptake varies by origin (Kil et al 2018;Mussino and Duvander 2016), we controlled for couples' origin, distinguishing couples where (i) both partners had a Swedish/Belgian origin, (ii) both partners had a migrant origin, (iii) the father had a migrant origin, or (iv) the mother had a migrant origin. We also controlled for region of residence (Nuts1) because, at least in Belgium-family policies, policy coverage and labour market conditions differ between regions.…”