Citizenship, work, and labor policy preferences

ABSTRACT

Although labor policy shapes and molds working conditions and the treatment of workers, and non-citizen Latinos are disproportionately exposed to poor working conditions, research on the relationship between the two – labor policy attitudes of non-citizen Latinos – remains scarce. This study draws on a dataset with an atypically large sample of both citizens and non-citizens and key indicators of labor policy to examine views of labor policy. The findings indicate that non-citizen Latinos are strongly supportive of labor policy that would make it easier to form labor unions and hold positive views of labor unions. The results extend our understanding of labor policy and the relationship between citizenship in the US. and support for labor unions – institutions that strengthen workers’ rights, reduce inequality, and improve working conditions.

KEYWORDS:

• Immigration
• political behavior
• public policy
• public opinion
• labor movements

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Although there are important differences among Latinos as a group, such as country of origin or history, NAES asked for respondents’ origin and offered only four options–Mexican (54% chose this option), Cuban (5%), Puerto Rican (3%), or other (28%, open-ended). Given the limited data, the analysis here employs Latino as a single category. Additional analyses that exclude Cubans from the sample do not substantively change the results and can be found in the SI. Additionally, the NAES data does not contain documentation data for non-citizens and therefore, although certainly important, there is no way to distinguish non-citizens with legal documentation from those without legal documentation.

2 Given the nature of the outcome variable in the favorability models – a 0–10 ordinal variable – ordered logistic regression may be used to specify the model instead of OLS. Additional models using ordered logistic regression models can be found in the SI. The results are substantively unchanged from those found here. Therefore, OLS is used here as it lends itself to more intuituive interpretation.

3 In line with NAES guidelines, the models are unweighted. The models include demographic controls which render weights unnecessary (Winneg, Kenski, and Adasiewicz 2006, 22).

4 In additional analyses, interactions between non-citizen and Black and non-citizen and Asian are not significant. While not the focus of this study, future work should examine additional subgroups. Results from these models can be found in the SI.

5 The median respondent is a 47-year-old male who is employed, married, not from a union household, with some college but no degree and with a household income of between $35,000 and $50,000, and moderate political ideology with independent party affiliation.