Electronic institutions are a formalism to define and analyse protocols among agents with a view to achieving global and individual goals. In this paper we propose a definition of norms for electronic institutions and investigate how these norms can be employed for verification and analysis. We offer automatic means to perform the extraction of subparts of an electronic institution in which norms hold true or can safely be avoided. These sub-parts can be used to synthesise norm-aware agents that will pursue or avoid commitments to norms.