Scholarship on reputations in IR has thus far left out one of the most striking features of
human psychology: identity. Categorizing others as either “us” or “them” is an automatic
and pervasive process that has significant implications for how reputations are generated and
maintained. We provide a theoretical framework—based on social identity theory—to explain
how ingroup bias affects both how we perceive other actors’ “type” and estimate their likely
behavior; in short, their reputations. Empirically, we provide two contributions. First, we
field a descriptive survey on one public and two elite samples to learn about the markers of
international ingroups and aid in the design of our experiment. Second, we pilot and field a
pre-registered experiment that tests implications of our identity-based theory of reputations.
We find that ingroup membership confers significant benefits to (1) reputations of all types
(financial, signaling, humanitarian and resolve) and (2) observers’ willingness to engage with
the state. We find little evidence that ingroup bias affects how past behavior is integrated into
reputations, and use follow-up studies alongside this null to help refine the scope conditions of
our theory.