close× Call Us
close×
Title Identifying Americans’ Immigration Policy Preferences via Conjoint Analysis
Post date 02/13/2019
C1 Background and Explanation of Rationale

Immigration is becoming a more divisive and contentious topic in American politics. While Republicans and Democrats were equally likely to support decreasing legal immigration to the U.S. in 2006, Republicans are now twice as likely as Democrats to favor decreasing legal immigration (Pew Research Center 2018). Additionally, a recent poll shows that 30 percent of Americans support President Trump’s immigration policies, 38 percent support the immigration policies of Democrats in Congress, and 24 percent support neither (The Washington Post 2018). The partisan division on the issue also increased over the past decade. While 49 percent of Republicans and Republican-leaning independents feel “unsympathetic” toward undocumented immigrants, only 13 percent of Democrats and Democrat-leaning independents harbor the same attitude (Pew Research Center 2018).

Despite the politicization of this topic, Hainmueller and Hopkins (2015) find evidence of a consensus on the type of immigrants American prefer. It is unclear, however, whether they have a similar consensual view on immigration policies. To address this, we conduct a conjoint survey experiment (Hainmueller, Hopkins, and Yamamoto 2014) with immigra- tion policy issues currently debated in American politics as attribute-levels. In addition, we include a component of randomized framing treatment in the survey to examine the effects of framing on immigration policy preferences. The survey experiment and its associated analyses provide an understanding as to what form of immigration policy individuals would support, what attributes are salient in this decision, whether framing affects this level of support, and whether preference and the effect of framing differ by the attributes of respondents.

C2 What are the hypotheses to be tested?

We aim to test three hypotheses and five research questions (RQs) through a conjoint analysis survey experiment with a randomized assignment of framing treatments.
Hypothesis 1 For each immigration policy issue, respondents have stronger preferences for policies that are more favorable toward immigrants than those that are less favorable.
Hypothesis 2 Across all immigration policy issues, border detainment is the most salient for respondents.
Hypothesis 3 Compared to the case in which “unauthorized immigrants” is used, respon- dents have stronger preferences for policies unfavorable to immigrants when “illegal aliens” is used and stronger preferences for policies favorable to immigrants when “undocumented immigrants” is used.
RQ 1 How do preferences for immigration policies differ depending on respondent charac- teristics?
RQ 2 How does framing change immigration policy preferences depending on respondent characteristics?
RQ 3 How do preferences (and differences in preferences) differ depending on how prefer- ence is quantified?
RQ 4 How do preferences for a specific attribute level change depending on the other at- tribute levels included in the proposal?
RQ 5 How do expectations about perceived crime rates and education attainment change depending on framing treatment and respondent attribute? – How much do respondents respond to framing treatments depending on their own attributes?

C3 How will these hypotheses be tested? * To test the hypotheses and research questions, we conduct a conjoint analysis survey exper- iment with randomized treatment assignments. We conduct group and sub-group analyses. See the attached for the survey design and plan of statistical analysis.
C4 Country United States
C5 Scale (# of Units) 1,350
C6 Was a power analysis conducted prior to data collection? No
C7 Has this research received Insitutional Review Board (IRB) or ethics committee approval? Yes
C8 IRB Number STUDY00031323, MOD00008205; MOD00008484
C9 Date of IRB Approval 10/19/2018; 11/27/2018; 1/23/2019
C10 Will the intervention be implemented by the researcher or a third party? Researchers
C11 Did any of the research team receive remuneration from the implementing agency for taking part in this research? No
C12 If relevant, is there an advance agreement with the implementation group that all results can be published? not provided by authors
C13 JEL Classification(s) D70, D80, F22, J68, C90