The Role of Geographic Context in the Local Politics of US Immigration

Abstract

In recent years, immigration has emerged as a highly contentious local issue in the United States, particularly in non-traditional settlement areas like new metropolitan gateways and suburban communities. This paper uses survey data from the Pew Research Center to analyse how geographic context within and across metropolitan gateways influences both perceptions of the local impacts of immigration, as well as respondents' attitudes toward immigration as a national issue. While immigrant concentration is positively associated with perceptions of immigration as a local problem, a relative absence of immigrants in a respondent's community predicts negative attitudes towards immigrants at the national level. Further, areas that respond unfavourably to immigration at the local level do not always coincide with the areas that are most opposed to immigration to the United States more generally. As such, this analysis identifies tangible differences between immigration as a local and national issue, suggesting a distinct unevenness in the multi-scalar politics of US immigration.

Keywords:

• Immigration
• Political attitudes
• Scale
• Contact hypothesis

Acknowledgments

I would like to thank Ross Macmillan for guidance with this project, and the two anonymous reviewers for helpful feedback on this article. This project was in part supported by NSF Doctoral Dissertation Research Improvement Grant BCS-0902685.

Notes

[1] For the cultural question, 0 = ‘The growing number of newcomers from other countries strengthens American society’ and 1 = ‘The growing number of newcomers from other countries threaten traditional American customs and values’. For the economic question, 0 = ‘Immigrants today strengthen our country because of their hard work and talents’, and 1 = ‘Immigrants today are a burden on our country because they take our jobs, housing and health care’.

[2] Missing values imputed.

[3] Following convention in the literature, the ICC for a multilevel model with a binary response variable is computed as τ₀₀ / (τ₀₀ + π²/3), where τ₀₀ = the between-aggregation unit variance and the within-aggregation unit variance is fixed at π²/3 (Snijders and Bosker 1999; Cho 2003).

[4] See note [1] above.