What's the Impact of Harsh Language on Immigration Preferences?

• In Event: Unpacking Immigration and Anti-Immigrant Attitudes

Abstract

This study explicitly tests the extent to which language used by political elites referring to immigrants influences Americans’ preferences on immigration policies and whether subsequent (bi)partisan cues induce a (less) more polarized response. Using three survey experiments on two separate samples – 2,053 from the non-representative crowdsourcing platform MTurk and 2,231 from NORC’s nationally representative AmeriSpeak panel - I provide new insights on how partisans evaluate the parties’ current welcoming/restrictive messaging strategies on immigration and counter their respective parties’ current positioning on the issue by including actual frames from credible political elites - former Presidents Ronald Reagan (welcoming) and Bill Clinton (restrictive).

Scholars argue that politicians can snatch voters from party rivals by signalling a tough position on immigration. But this can be risky because it may have lasting effects on how the voters define the party crafting these messages, and alienate much larger portions of the electorate. Although elected officials across the political spectrum have a long history of framing immigration as an important policy problem, the negative immigrant-related messages introduced in recent years were significantly more likely to come from Republicans and experimental studies that exposed respondents to political messages on immigration found that negative frames about restrictionist policies only seemed to resonate among Republican voters. This growing body of evidence raises some important questions. First, in this current climate – where immigration is highly salient and highly polarized – how much influence do negative frames have on shaping individuals’ views on immigration policy if there is no political information tied to the message? Second, if individuals are affected by the prolonged exposure to their party’s current immigration frames – restrictionist for Republicans and welcoming for Democrats – then how would partisans react to a contrasting frame from a credible in-party elite? Third, do party sponsor cues trigger more polarized responses among partisans for the policies displayed or will the party faithful stick to their party’s current position on immigration? And finally, how do partisans respond to information that shows both parties have supported a particular immigration policy that runs counter to the current party rhetoric on the topic?

To determine the extent to which harsh language on immigration negatively influences policy views the MTurk sample (N=2,053) was split into four groups. First, ¼ of respondents were presented with a restrictive Clinton frame and a ¼ of respondents presented with a welcoming Reagan frame, but neither group received any information regarding the origins of the statement. To test how partisans react to a frame that contrasts the current parties’ position on immigration the remaining half of the sample received the same welcoming (¼) and restrictive (¼) frames that also included the name and party of the political elite who made the statement. Respondents were then asked their views on the current levels of immigration. The harsh language test using the same welcoming and restrictive frames (with attribution only) was replicated on the nationally representative AmeriSpeak sample (N=2,231) and followed with a second level treatment testing the effects of (bi)partisan cues on two policy options each supported by both parties - in-state tuition for unauthorized immigrants (positive) and denaturalization of citizens (negative).

Although, party identification is the greatest predictor of preferences, the pattern of results indicates that Americans respond to the frames with remarkable uniformity. On the first test Democrats and Republicans alike were more welcoming when they were presented with the Reagan frame and more punitive when allocated to the Clinton frame, a consistent finding across both samples. The test of attribution put to the non-representative sample failed to shift the views of Republicans or Democrats. Similarly, the nationally representative sample remained unmoved by the bipartisan cues in the second experiment, but somewhat unexpectedly Republicans and Democrats allocated to the Democratic cues consistently provided more negative responses (significance varies across items). Additional analysis examining the potential spillover effects revealed that Reagan’s welcoming frame provoked more positive responses among Democrats on the positive in-state tuition policy, but not on the denaturalization scenario. The results were less consistent among the Reagan-frame Republicans. Only those who were allocated to the bipartisan cue were significantly more positive regarding the in-state tuition scenario, but this effect did not hold on the denaturalization scenario. But Reagan-frame Republicans who were allocated to the Republican cue were significantly less negative in their response to the denaturalization cue.

Author

• Melissa Shannon, London School of Economics