Sibling spillovers and the choice to get vaccinated: Evidence from a regression discontinuity design.

We investigate the effects of introducing population-wide free-of-charge Human Papillomavirus (HPV) vaccination programs on the targeted adolescent cohorts and their siblings. For identification, we rely on regression discontinuity designs and high-quality Danish administrative data to exploit that date of birth determines program eligibility. We find that the programs increased the HPV vaccine take-up of both the targeted children (53.2 percentage points for girls and 36.0 percentage points for boys) and their older same-sex siblings (4.5 percentage points for sisters and 3.5 percentage points for brothers). We show that while the direct effects of the programs reduced HPV vaccine take-up inequality, the spillover effects, in contrast, contributed to an increase in vaccine take-up inequality highlighting the potential importance of spillover effects in the determination of distributional consequences of public health programs. Finally, we find some evidence of cross-vaccine spillovers.