The social roots of Asian American partisan attitudes

ABSTRACT

An important question in the study of Asian American political behavior is why members of this group, who have some Republican predispositions, are strong Democratic supporters. How do Asian Americans develop preferences for the Democratic Party? This paper considers this question through a theory-building case study of political socialization in Houston, Texas, a mixed-partisan area where many Asian Americans support Democrats. Using qualitative interviews and supplemental survey data, I examine voting behavior and partisan identification, developing a socially based explanation for high levels of Democratic support. I argue that Asian Americans develop Democratic preferences through interactions within peer groups. These peer groups vary in composition by generational status, leading to different reasons for Democratic support across generations. I find that first-generation Asian Americans primarily interact with fellow Asian immigrants. Relative to their children’s generation, they develop positive attitudes about conservative ideology but vote for Democrats because they perceive contemporary Republicans as ideologically extreme. In contrast, transitional-generation and second-generation Asian Americans have racially diverse peer groups. Rarely discussing politics at home, they develop pro-Democratic attitudes through interactions with liberal friends in educational settings.

KEYWORDS:

• Asian American politics
• political socialization
• political parties
• immigrant political behavior
• generational status

Acknowledgements

I would like to thank Christopher Achen, Martin Gilens, Tali Mendelberg, and Dara Strolovitch for their helpful feedback on this work. I also thank the Asian American residents of Houston who participated in this study, Anne Chao and Lisa Gray for their help with subject recruitment, and the Princeton University Program in American Studies for funding this work.

Disclosure statement

No potential conflict of interest was reported by the author.

ORCID

Tanika Raychaudhuri http://orcid.org/0000-0002-1927-3768

Notes

1 Exit poll data is unreliable for geographically concentrated minority groups, but useful for understanding how Asian American vote choice has changed over time (Barreto et al. 2006).

2 “First-generation” refers to immigrants who arrived in the US at 21 or older. “Transitional-generation” refers to immigrants who arrived at 21 or younger. “Second-generation” refers to those born in the US, with at least one parent born outside of the US.

3 While this research focuses on Asian Americans, I expect that a similar process may also apply to other immigrant groups, like Latinos.

4 The interview protocol is available in the online appendix (Appendix 17).

5 Population parameters are from the 2016 American Community Survey. National origin statistics are from the 2010 US Census.

6 While non-citizens make up 14% of the sample, they are relevant to the study because they make up 36% of the Houston Asian American population. Also, experiences prior to naturalization are relevant to the political socialization of foreign-born citizens.

7 Independents are included in the “non-partisan” category because they are not strongly attached to either party.

8 This analysis was conducted separately by national origin when possible, for subgroups large enough to conduct a sufficiently powered regression analysis (Tables A6 and A7, Appendix 13–14). Although many of these findings are robust to the main results, there are some differences across national origin groups, which are discussed in the online appendix. The results should be interpreted with caution due to the small sample sizes.

9 I conduct a mediation analysis to test whether interracial friendships have an intervening effect on the relationship between generational status and having any partisan identification. There is no significant association between generational status and Democratic identification, so a mediation test is not appropriate. See Table A8 (Appendix, 15–16) in the online appendix for the results. The results do not find support for mediation.

Additional information

Funding

This work was supported by Princeton University.