Attitudes about refugees and immigrants arriving in the United States: a conjoint experiment

ABSTRACT

From 2013 to 2018, the global refugee population increased by about 50 per cent. Although public opinion toward refugees plays a key role in shaping related policies, scholarly research on this topic remains limited. Using original data from a nationally representative conjoint survey experiment conducted in 2019, we examine whether US citizens’ attitudes toward refugees are distinct from their attitudes toward other types of immigrants to the US, and how applicant attributes shape American citizens’ preferences about admission of foreigners. We find that immigrant and refugee reasons (as defined by law) for migrating to the US affect attitudes, with refugee reasons garnering greater support. The labels “immigrant” and “refugee” on their own have smaller effects, in general, but can be salient in combination with other characteristics. We also find marked effects of newcomers’ characteristics, especially country of origin, profession, gender and religion.

KEYWORDS:

• Refugees
• immigration
• public opinion
• attitudes
• conjoint methods
• conjoint experiment

Acknowledgements

The authors wish to thank Lina Newton, Marcus Johnson, Jennifer Laird, Ke Li, James Rodriguez and Min Zhu for invaluable feedback on earlier drafts of this paper. Mariah Lutchman and Aida Murati, John Jay College McNair Scholars, provided research assistance on this project.

Disclosure statement

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

IRB approval

This study has been granted an exemption by the CUNY IRB (protocol number 2019-0526) and the Syracuse University IRB (protocol number 19-209).

Notes

1 We use the terms “newcomers”, “foreigners”, or “foreign newcomers” to refer to all non-citizens arriving in a country with the intent to stay.

2 By “attitudes”, we refer to support or lack thereof for admission of foreign newcomers.

3 As we explain below, the US distinguishes between asylum-seekers (who are physically present in the country) and refugees (who are physically located outside the country). We focus on refugees in this study because the government controls which and how many of them will be admitted (i.e. resettled) from abroad.

4 The definition of “refugee” under US law, as codified in the 1980 Refugee Act, is based on the United Nations’ 1951 Refugee Convention. However, the US espouses a unique definition of the terms “refugee” and “asylee”. Both refugees and asylees are people “who are unable or unwilling to return to their country of origin or nationality because of persecution or a well-founded fear of persecution”. The primary difference between a refugee and an asylee in the US is a procedural one (Bray 2016): “refugees are physically located outside the country while asylum seekers are already present within US borders at the time of application” (USCIS 2015).

5 The International Organization for Migration (IOM) deals mostly with service provision. Meanwhile, the UN Convention on Migrant Workers remains undersubscribed (particularly by destination countries), as are the conventions approved by the International Labor Organization (ILO).

6 A JavaScript coding error in the soft launch of the survey experiment caused the loss of 186 ratings. Hence, the total number of ratings is 5,114 instead of 5,300.

7 Nonparametric methods do not make explicit assumptions about the functional form of the function f, which determines the relationship between the outcome and the input (e.g. James et al. 2013).

8 While the randomization of profiles leads to “bogus refugees” being overrepresented in the data, many Americans misunderstand who refugees are (e.g. Ipsos 2016; Thorson and Abdelaaty 2022) so we were very interested in capturing the effects of such combinations.

9 Alternatively, while other studies such as Bansak, Hainmueller, and Hangartner (2016) had specified specific types of persecution – political or religious – we attempted to limit variable levels in our study by just using the term “persecution” on its own. However, it is possible that the lack of the political or religious modifier prevented some respondents from understanding the meaning of the term.

Additional information

Funding

This research was funded by two awards, Cycle 50 PSC-CUNY Research Award TRADB-50-315 and a John Jay Office for the Advancement of Research Seed Money Award 90616-24-01.