In my dissertation, I develop a novel cognitive-psychological framework for the study of public attitudes toward immigration. In the first chapter, I lay out the general approach to studying political cognition based on the associative view of human memory. I argue that individuals imagine abstract political categories, such as “immigrants,” by connecting them to more concrete social and demographic attributes. These images, not the underlying social reality, then powerfully define policy preferences. I also argue in favor of using implicit measurement techniques for measuring images of political objects, such as the implicit association test. Using original survey studies on gendered images of political parties and the racialization of poverty, I document both promise and limitations of timed-response tasks as measures of cognitive associations in politics.

In other chapters of my dissertation, I apply this theoretical approach to explain the dynamics of public attitudes toward immigration in the United States and Britain. In the second chapter, I address an essential controversy in existing literature concerning the importance of racial animosity in public opposition to immigration. Recent experimental studies suggest that, contrary to massive observational evidence, distinctively racial cues play a limited role in driving anti-immigration attitudes. I argue that the degree to which natives think of immigrants in racial terms is an integral part of opposition to immigration in developed democracies, but survey experiments to date fail to capture that key variance. Using two original survey studies, I demonstrate that variation in the racialization of immigrant groups measured both implicitly and explicitly is strongly associated with group-specific prejudice, attitudes toward immigration, and partisan affect.

In the third dissertation chapter, I study broader beliefs about immigration and immigrants. In two original survey studies, I assess respondents’ perceptions about immigrant population on a number of important dimensions: origin, education, skills, language proficiency, dependency on welfare, and respect for laws. I demonstrate that supporters of the major national parties significantly diverge in terms of their beliefs about immigrants. I show that these beliefs strongly predict attitudes toward immigration above and beyond partisanship, ideology, and generalized ethnocentrism. I also match survey and aggregate data to investigate the local roots of beliefs about immigration and find that ethnic composition of respondents’ localities affects beliefs about the ethnic composition of the immigrant population.