ABSTRACT
How do racial and gender stereotypes affect immigration attitudes? Studies on this topic have focused primarily on developed rather than developing countries. To close the gap, we explore the phenomenon in China, a developing economy that is evolving from a source of emigration to a destination of immigration with its citizens having little direct experience with immigrants. Via two waves of survey experiments with over 4,000 respondents, we find that gender cues of female immigrants increase public approval of immigration and liberal immigration policies, while racial cues of Black immigrants discourage them. Further analysis reveals that gender bias is likely the product of sociopsychological concerns over cultural threats and that racial bias stems from concerns over public safety. Our findings contribute to the literature on the racial stratification and discrimination of immigrants in the context of developing countries.
Acknowledgements
The authors thank the editors and two anonymous reviewers for their comments and feedbacks. The authors appreciate suggestions from Efrén O. Pérez, Stella M. Rouse, and panelists in 2020 SPSA annual conference. The authors are grateful of excellent research assistance from Yishan Jin, Jiazheng Lu, Hao Yang, and Yilun Yao. Parts of this project are generously supported by Zhejiang University Fund for Science Research Development (S20210156) and New Faculty Start-Up Grant at City University of Hong Kong (7200670). All errors are our own.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
3 http://www.gov.cn/xinwen/2018-03/24/content_5277006.htm; see also https://www.scmp.com/news/China/society/article/3019034/China-relaxes-immigration-rules-attract-and-retain-more-highly.
5 In Chinese, Xiao Xie means ‘the young person whose surname is Xie’ with no gender indicated. In the vignette, it could mean either a young man or a young woman, depending on the context.
6 In Mandarin, skin tone is widely used to identify race. We used skin tone to make the experiment more understandable.
7 Images as treatments may bring unintended confounding factors. In the Appendix, we have more details to discuss how we minimize the potential confounding factors and description of the treatment materials.
8 In the Chinese version. The characters of their Chinese names varied to coordinate the gender treatment in the image while the pronunciation of their names kept the same.
9 All questions had a four-point scale and were recoded to [0-1] in the analysis, unless specified.
10 We designed this question wording in case that the respondent was married. In Chinese version, the question is clear that the respondent only needs to answer the one that fits their conditions.
11 The descriptive statistics, balance check and full regression results are available in the Appendix.
12 The descriptive statistics, balance check and full regression results are available in the Appendix.