Justification within and between social contexts with the possibility of choice deferral

Abstract

This paper provides axiomatic characterizations for a class of models where a decision maker (henceforth, DM) maximizes a single preference ranking subject to a shortlist that depends on the underlying choice set and social context. Thus, the primitive is a dataset of choices from every choice set – social context pair that is called the context. The DM is supposed to choose the most preferred option from the shortlist of each context. However, due to existence of conflicting norms (or norms uncertainty), he/she might choose no option in some contexts. Therefore, unlike existing justification models, choice deferral is possible in our framework.