The argument of this article has three parts: first, we argue that the theory of social norms is relevant to the study of corruption. This is so because corruption is a phenomenon that occurs in a specific context, within a given regulatory system, which allows certain interactions to be defined as "corrupt". Second, we show that Coleman's theory of social norms suffers from a theoretical deficit, since it does not admit certain subtleties in the explanation of the macro-to-micro transition (that is, how macrosocial phenomena influence individuals). Based on Coleman's scheme of levels of social interaction, we point out an asymmetry not theorized by Coleman between the mechanisms of the micro-to-macro passage, which generate a macro level that is objective or, at least, common to many, and the macro-to-micro mechanisms, which start from a macro level that is not necessarily perceived in the same way by individuals, but is mediated by their beliefs about society; and this is more noticeable when it comes to a phenomenon such as corruption, in which there are strong incentives for secrecy, so it is not possible to affirm that the social level is objective (in terms of its perception by individuals) or common to many. There are, as we will see, mechanisms that intervene in the explanation of corruption that are related to the influence of social norms. Third, we will exemplify the macro-to-micro transition through the notion of normative expectations, and study the mechanisms that make it easier for individuals to perceive corrupt behavior as legitimate. Specifically, we will review different mechanisms that are interrelated and that refer to: modes of transaction, cognitive dissonance, narratives, and, pluralistic ignorance.