Today a common goal in the area of email security is to provide protection from a wide variety of threats by being more predictive instead of reactive and to identify legitimate messages in addition to illegitimate messages. There has been previous work in the area of email reputation systems that can accomplish these broader goals by collecting, analyzing, and distributing email entities' past behavior characteristics. In this paper, we provide taxonomy that examines the required properties of email reputation systems, identifies the range of approaches, and surveys previous work.