Unintended health consequences of Swedish parental leave policy (ParLeHealth): protocol for a quasi-experimental study

Introduction Sweden has long been praised for a generous parental leave policy oriented towards facilitating a gender-equitable approach to work and parenting. Yet certain aspects of Swedish parental leave could also be responsible for the maintenance of (or even the increase in) health inequalities. Using a 'Health in All Policies' lens, this research project aims to assess the unintended health consequences of various components of Sweden's parental leave policy, including eligibility for and uptake of earnings based benefits. Methods and analysis We will use individual-level data from multiple Swedish registers. Sociodemographic information, including parental leave use, will be retrieved from the total population register, Longitudinal Integration Database for Health Insurance and Labour Market Studies and Social Insurance Agency registers. Health information for parents and children will be retrieved from the patient, prescribed drug, cause of death, medical birth and children's health registers. We will evaluate parents' mental, mothers' reproductive and children's general health outcomes in relation to several policy reforms aiming to protect parental leave benefits in short birth spacing (the speed premium) and to promote father's uptake (the father's quota) and sharing of parental leave days (the double days reform). We will also examine effects of increases in basic parental leave benefit levels. Using quasi-experimental designs, we will compare health outcomes across these reforms and eligibility thresholds with interrupted time series, difference-in-difference and regression discontinuity approaches to reduce the risk of health selection and assess causality in the link between parental leave use and health. Ethics and dissemination This project has been granted all necessary ethical permissions from the Stockholm Regional Ethical Review Board (Dnr 2019-04913) for accessing and analysing deidentified data. The final outputs will primarily be disseminated as scientific articles published in open-access, high-impact peer-reviewed international journals, as well as press releases and policy briefs.