ABSTRACT
The number of individuals currently living in a country other than the one in which they were born peaked at more than 244 million in 2015. This makes the global society a culturally and racially diverse one. Although unpopular with some residents, immigrants are the reality of contemporary society. This study discusses immigration as a global phenomenon and the creation of immigrant-friendly society as a necessity. The study utilizes data from the 2014 General Social Survey on American attitudes toward immigrants to create the feelings toward immigrants (FTI) index and investigates the determinants of the feelings toward immigrants in the United States. The study finds that college education, income, political views, citizens, and residing in the West and the Northeast United States affect feelings toward immigrants. Interestingly, the study reveals that college education enhances the positive perception of liberals and moderates the negative views that conservatives and citizens might have. The study concludes that policymakers seeking to create immigrant-friendly society should focus on the above factors rather than race, attendance, and participation at religious events and other variables that were not significant.
Notes
1 For example, the corporate tax rate in China is 25% whereas in Mexico it is 30%; both are much less than the 40% in the United States (https://home.kpmg.com/xx/en/home/services/tax/tax-tools-and-resources/tax-rates-online/corporate-tax-rates-table.html. Accessed on December 12, 2016).
2 Variance Inflation Factor [VIF] scores for models 1 and 2 are generally below 4. VIF scores are considered problematic only when they are above 4, hence no harmful multicollinearity is present in both models. However, there are two high VIF scores reported in model 2 as a result of the interactions with college and citizen [College = 21 and CollegeXCitizen = 19]. These do not pose problems in the model because this type of multicollinearity would not change the coefficients of the regression and the p-values are expected to remain unchanged even if it is altered to remove the multicollinearity (Friedrich, Citation1982; Jaccard & Turrisi, Citation2003).