Human papillomavirus vaccine and risky sexual behavior: Regression discontinuity design evidence from Brazil

This study aims to analyze the hypothesis of moral hazard caused by vaccination against human papillomavirus (HPV), regarding girls' beginning of sex life and, once they have begun their sex life, to understand whether it reduces the probability of girls using a condom in their first sexual intercourse. The data are from the 2015 National Survey of School Health (PeNSE). The model used is the regression discontinuity, with the discontinuity in the age of the girls who were able to participate in the public vaccination campaign. The results of the so-called Fuzzy-RDD show that the campaign was effective in increasing the likelihood of vaccination by 26.7-27.6 percentage points. There is no observable effect on young women initiating their sex lives or refraining from using condoms. The results are tested by several robustness methods. This is the first work to use quasi-experimental models in a developing country with low vaccination coverage in Latin America and show that it is necessary to increase awareness campaigns with parents, so they will know that there is no effect of vaccination on the beginning of sex life and condom use.