Chuang Chen

Do immigrant presence and racial segregation affect different individuals’ attitudes toward immigration? There is a controversy in previous literature about whether more immigrants lead to more or less anti-immigration attitudes. I argue that the disagreement is because they ignore the degree to which the immigrants (and other minorities) are segregated. In the U.S. context, I hypothesize that for conservatives, higher immigrant/racial segregation is related to more pro-immigration attitudes because a lack of intergroup communication functions as a “shield” against more salient group membership and substantial prejudice against the outgroups. With data covering five presidential elections in 2008–2024 from the U.S. Census, the American Community Survey, and the Cooperative Election Study, I use Bayesian multilevel linear regressions with immigration attitudes as the DV. Most results support the hypothesis. This paper contributes to immigration politics by comparing the different effects of immigrant presence and racial segregation interacting with ideology and providing a modified version of the threat theory in explaining the divergent effects of racial segregation on liberals’ and conservatives’ immigration attitudes.

In survey research on political attitudes, a question that is often left unanswered is the question of ideology, especially in Latin America. What does it mean when they claim to have no ideology? What motivates them to place or refrain from placing themselves on the ideological scale? In Chile, where the response rate for the ideology question is very low, reasons for this phenomenon, as well as its changes, have seldom been explored. This study proposes the concept of “immigrationalized” ideology and hypothesizes that the attitude on the immigration issue, which has not been politically salient until recent presidential elections, could drive voters to return to or withdraw from the ideological scale. Specifically, a strongly negative attitude toward immigrants could drive those who used to be rightists to exit ideology in the new election, while the same attitude could drive those who used to claim to have no ideology to have an ideology in the new election. With panel data from Estudio Longitudinal Social de Chile (2016–2023), I run Bayesian multinomial logistic regressions with comuna-level random effects. This study contributes to research on immigration attitude and ideology in Latin America by analyzing the group of respondents who have often been left out in previous studies but have been playing a nonnegligible role in elections.

Does the integration of immigrants ease prejudice against them? While the short-term increase in immigrants may trigger anti-immigration attitudes among natives, many scholars argue that the contact between immigrants and non-immigrants will moderate these attitudes across time. However, it remains unclear why the negative attitudes toward some groups of immigrants do not improve over time. I propose the “integration backlash theory” hypothesizing that increasing economic integration of immigrants will strengthen the negative attitudes and the support for the radical right party among conservative non-immigrants, while cultural integration dampens it. Furthermore, this effect is stronger if natives’ economic condition has worsened than if it has improved. With data from Statistics Sweden, the Swedish National Agency for Education, and the SOM Institute Cumulative Dataset, I run multilevel linear regressions. Most findings support my hypotheses. This research contributes to the study of immigration politics by showing that the effects of economic and cultural integration of immigrants on immigration attitudes and support for the radical right are different and are conditioned by the local economic situation of non-immigrants.

My dissertation adopts various advanced statistical methods. After matching fine-grained census data at the county level (or even lower) with individual-level surveys, I use Bayesian multilevel linear regressions and binomial/multinomial logistic regressions to analyze the data, combined with techniques of geospatial analysis. The local demographic data provides information about immigrants that has been overwhelmingly ignored by previous scholars (such as segregation), and the rigorous methods help isolate effects of strong confounders. I also use Python to scrape electoral results of several countries online for macro-level analysis.