People near each other in a social network tend to act similarly; you can predict what one of them will do from seeing what their neighbors do.  Is this because they are influenced by their neighbors' actions, or because social ties tend to form between people who are already similar, and so act alike, or some of both?  We show that observational data generally can't answer this question, unless accompanied by very strong assumptions, like measuring everything that leads people to form social ties.  Most observational studies thus provide no evidence at all about the existence or strength of social influence.  There are, however, some situations where the global configuration of the social network can tell us enough about its individual nodes to get around this.