‘Bad hombres’? An examination of identities in U.S. media coverage of immigration

ABSTRACT

When discussing immigration on the campaign trail in 2016, Republican presidential nominee Donald Trump used the term ‘bad hombres’ to criticise immigrants, and in particular Latino men. Using a novel dataset of images accompanying immigration news stories in three major news magazines from 2000–2010, we examine the extent to which press coverage helped set the stage for this sort of racialised and gendered association. We find overwhelming support for the ‘bad hombres’ hypothesis, demonstrating that in comparison to the true demographics of immigrants in the U.S., press portrayals significantly over-represent Latino men and under-represent all other immigrant groups. Our results are particularly important in light of research demonstrating that individuals’ political attitudes are influenced by the perceived ‘target group,’ or beneficiary, of a given policy, as well as scholarship demonstrating the media’s role in fostering a sense of ‘Latino threat.’

KEYWORDS:

• Immigration
• media coverage
• Latino
• framing
• intersectionality
• images

Acknowledgements

We are incredily grateful to our Research Assistants, Alyssa Clark, Kallie McLaughlin, Lauren Meininger, Stephanie Shady, and Chanel Vismara. Early versions of this paper were presented at the annual meetings of the American Political Science Association and the New England Political Science Association, and we appreciate the helpful feedback from discussants and panelists.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 As defined by Chavez (2008, 24), the term racialised refers not to ‘genetic-based categories of race but, rather, labels that are socially and culturally constructed based on perceived innate or biological differences and imbued with meanings about relative social worth.’

2 Our expectation that Latinos will be significantly overrepresented in press coverage of immigrants contrasts with research demonstrating that in general, Latinos are underrepresented in media portrayals (van Doorn 2015; Clawson and Trice 2000; Wilkes and Valencia 1989; Bowen and Schmid 1997). For instance, in news stories about poverty, Latinas are particularly absent; while Latinas constitute 15.7 percent of the actual poor, they are featured in just 3.8 percent of news images published between 1992 and 2010 (van Doorn and Bos 2017).

3 In addition to the 2006 protests, we also excluded images from the 2000 controversy that ensued when five year old Elián González was rescued at sea after his mother drowned in her efforts to bring him to the U.S. from Cuba. Protests unfolded across Miami in an effort to keep him in the U.S. before he was finally returned to his father to live in Cuba. Due to the difficulty of evaluating who in these situations is an immigrant, we did not instruct our RAs to code these images.

4 Due to limited data availability, our analysis of U.S. News and World Report extends only through December 31, 2009. Our specific focus on ‘immigrant’ and ‘immigration’ is consistent with Pérez (2016). Although some of our stories included discussion of ‘migrants’ or ‘refugees,’ we did not explicitly search for these terms. Focusing on a similar time period, Gabrielatos and Baker (2008) find that in the United Kingdom, the news media described all of these groups similarly; more recently, Blinder and Allen (2016) find that between 2010–2012, the British press continued to describe migrants and asylum seekers similarly to immigrants, but portrayals of refugees began to diverge. Future research could extend this analysis into the U.S. context.

5 We eliminated cartoons from our sample given the small number of images to fall into this category. We also eliminated author photographs from our image count.

6 Notably, the 2005 ACS data is majority men in terms of its U.S. foreign-born sample, which is not in line with the overall trends during this time period.

Additional information

Funding

We are thankful to the APSA Fund for Latino Scholarship, as well as Clark University and Texas Christian University, for their support of our data collection and analysis.