A critical role of online platforms like Facebook, Wikipedia, YouTube, Amazon, Doordash, and Tinder is to moderate content. Interventions like banning users or deleting comments are carried out thousands of times daily and can potentially improve our online spaces. However, researchers, governments, and even platforms are ill-informed about the impact of content moderation.

In this thesis, we study the causal effect of various content moderation practices using observational studies and a large-scale field experiment. With online traces that range from users' daily activity on fringe websites to comments on online communities, we develop and apply observational and experimental study designs to understand how content moderation shapes subsequent user behavior and our online spaces.

The thesis is divided into two parts, each containing three studies. In the first part, we consider interventions targeting individuals or collectives in social media, conducting studies that assess the effect of banning influencers, online communities, and even entire online platforms. In the second part, we consider interventions targeting the creation of content in social media at different points in time. We study the effect of automatically deleting content after it has been posted, requiring content to be manually approved before posting, and helping users create content that is not rule-breaking.

Altogether, the scientific findings and methods presented advance our understanding of how content moderation shapes user behavior and online platforms and can inform the design of content moderation interventions, systems, and policies.