Social identities are fascinating psychological phenomena that explain a host of behaviors and attitudes. Scholars from multiple disciplines have examined how group-based attachments impact a variety of social, political, and economic outcomes. In this dissertation, I show where group-based attachments come from and why they matter for politics for Latinos and Asian Americans living in the U.S. I find that pan-ethnic and national origin attachments vary considerably given one's social and environmental context. I show that an individual's context is an essential factor in understanding the variation in attachment towards specific groups and the strength of that attachment. I develop a theory of context that connects group-based identities to one's local environmental and social contexts -- arguing that context provides a set of cues and stimuli which structure the self-categorization and self-stereotyping process, the psychological process where individuals shed ``individual-ness'' and opt for a ``group-ness'' in their self-concept.

I then show the conditions under which group-based attachments are consequential for engagement in the U.S. political system. I introduce and test the identity portfolio theory, a framework that shows how group-based identities that drive political outcomes are dependent on three politicization mechanisms: 1) the strength of attachment to a social identity category; 2) the strength of attachment to other social identity categories, and 3) the degree of permeability between the social identity categories held in one's identity portfolio. While recognizing that people are attached to multiple social identity categories, I show the conditions under which some of these identities are important for politics, but only at certain times or under certain conditions.

I conclude by discussing the implications of this work and outlining the contributions it makes to our understanding of American politics broadly defined.