Nowadays, social media became one of the most important channels for information consumption by individuals. Due to the power of the digital “word-of-mouth”, businesses increasingly use social networks for the spread of marketing content and political organizations spread opinions and views. Policy makers are concerned that, along with other content, extremist, aggressive or harassing messages go viral on social networks and quickly reach a wide audience.

We analyze the Network Enforcement Act (NetzDG) established by the German government in order to constrain the free dissemination of “Hate Speech” on major social networks in Germany.. The Network Enforcement Act was approved by the German Bundestag on 30th June, 2017. The law came into force on 1st October, 2017, and until the 1st January, 2018, major social networks were obliged to establish services that enable timely deletion of messages containing hate speech. In our study, we use the gradual introduction of the bill as a natural experiment for the causal identification of the effects of regulation on the sentiment of politicians’ online posts on Twitter. We study the effect of the regulation on the content posted on Twitter by German politicians and the subsequent effect on users’ engagement reacting to these tweets.