Latino Identities in Context: Ethnic Cues, Immigration, and the Politics of Shared Ethnicity

Abstract

This dissertation is a collection of three essays examining the relationship between immigrant political rhetoric and identity among Latinos in the United States. To achieve this task, this study uses empirical evidence from a national survey and original data collected from experiments in New York City and Los Angeles. The first essay identifies three forms of Latino identity most relevant to political decision-making: national origin, pan-ethnic, and American. I find that levels of acculturation as defined by immigrant status and English language strongly predict American identification. Latino identities inform support on immigrant issues. Latinos with higher perceptions of national origin and pan-ethnic interests are more pro-immigrant on issues pertaining to the rights of undocumented immigrants. The second essay investigates how exposure to explicit and implicit cues within anti-immigrant rhetoric shape the voting decisions of non-Mexican Latino groups in New York City. I test the effects of pan-ethnic, nationality-based, and counter-stereotypical political appeals on candidate support. I find that nationality-based appeals directly or indirectly targeting Mexican immigrants do not activate identity in vote choice, only explicit, pan-ethnic cues implicating all Latino immigrants activate "Latino" group interests in voting decisions. The third essay tests whether political processes of collective identity observed among non-Mexicans in New York City are generalizable to Mexican and non-Mexicans in Border States. Conversely, I find that only nationality-based political appeals targeting Mexicans activate Mexican group interests in vote choice. These results do not extend to non-Mexicans. Anti-immigrant messages did not activate identity in voting. Overall, these findings suggest that identity activation in the context of threat may work differently for Mexican and non-Mexican Latino groups in the United States.