This dissertation addresses three events that had an impact on violence against women. Violence could be represented in several ways: physically, psychologically, or sexually. In this work, the effect on the three forms of violence are explored. First, I study the increase in the number of rapes in the 24 subsequent hours after a major YouTube outage occurred on October 16th, 2018. Second, I explore the long-term causal impact of military conscription on sexist attitudes and intimate partner violence. Lastly, I analyze the extent to which the quarantine implemented following the coronavirus pandemic had unintended consequences on intimate partner violence (IPV). A feature of this dissertation is the accounting of motives and mechanisms, to the extent possible, associated with each event. First, using high-frequency crime data from the U.S., the study documents a significant increase in rapes in the 24-hour period following the major YouTube outage that occurred on October 16th, 2018. We investigate various potential underlying channels that may link the YouTube outage to the subsequent observed increase in rapes: we explore a direct effect on crime, time substitution, an effect on the consumption of drugs and alcohol, and the increase in pornography viewing. The overall evidence only supports the hypothesis that the increase in rapes could have been driven by an increase in pornography viewing. In my second essay, I provide empirical evidence on the long-term causal impact of military conscription on sexist attitudes and intimate partner violence. To address potential endogeneity, we exploit the conscription draft lottery in Argentina. We combine the draft administrative data with self-reported survey data. We find that conscripted men are more likely to report embracing more sexist attitudes in dimensions such as justification of sexism and violence, sexual machismo, negative attitude towards homosexuality, old-fashioned sexism, and hostile sexism. We also find that men who served are more likely to self-report engagement in intimate partner violence, as measured by non-physical abuse and physical violence. In my third essay, we use self-reported survey data from Argentina to study the extent to which the quarantine implemented following the coronavirus pandemic had unintended consequences on intimate partner violence (IPV). Since the disease arrived late to Argentina and the government reacted fast, the full national lockdown was imposed when few people felt threatened by the virus. The quarantine decree also established clear exceptions for subsets of the population and, for reasons plausibly exogenous to the prevalence of IPV, only some individuals were placed in quarantine. Exploiting this variability in exposure, we find a positive link between quarantine and IPV.