Partisanship in Everything: Public Opinion about Immigration

by John Sides on September 24, 2012 · 6 comments

in Immigration,Public opinion

Bill Bishop reports on a new poll of voters living in rural areas in the swing states of Colorado, Florida, Iowa, Nevada, New Hampshire, North Carolina, Ohio, Virginia and Wisconsin.  The poll contained the experiment depicted above, which confronted voters with two different positions on immigration, but varied whether those positions were attributed to the Democratic and Republican parties or had no party labels attached.  The result:

Republican-leaning rural voters supported the Republican position when it was labeled “Republican.” When the same position was asked without any partisan label, its support dropped 10 percentage points.
That’s the power of partisanship in today’s politics.


Darin Self September 24, 2012 at 11:22 pm

I actually have a working paper on anti-immigrant behavior. I found that it isn’t just relative number of immigrants to native population or unemployment that leads to political behavior against immigrants. I found that what is most importantly the rates of change in both immigration and unemployment together. I give a basic explanation why immigration isn’t a “hot button” issue right now.

David Karger September 25, 2012 at 2:24 am

I think there’s some subtlety required in interpreting this experiment. The quoted “positions” do not contain enough information to let them be interpreted out of context. They contain codewords like “rule of law”, “amnesty”, “comprehensive reform”, and “compliance with the law” whose meanings are ambiguous. While it’s possible that subjects are being swayed to go along with their party’s positions, another possibility is that the subject are sticking to their own preferences, but changing their interpretation of these position statements based on the (believed) context of who is using them.

Idiot September 25, 2012 at 10:29 am

Agreed. Also note that technically, both positions aren’t in conflict with one another. You could secure the rule of law while still forcing illegal immigrants to pay taxes and learn English.

nyetaryan September 25, 2012 at 2:09 pm

It was an experiment, and if the experimenter randomly assigned the labels to the statements, then all the confounding issues you guys mention would have been adjusted for. Taking John Sides at his word, the data reveal the independent influence of party label, period.

idiot September 25, 2012 at 2:36 pm

Well, yes, party labels do play a role; no one’s denying that.

We just wanted to add more information about the politics of immigration and about the question wording…because there’s nothing else to talk about.

albatross September 26, 2012 at 9:44 am

One intersting wrinkle here is that knowing the history of a speaker helps you parse his current statements. If Richard Dawkins responds to news of some calamity with “God help us,” it’s a very good bet he was using a figure of speech, whereas Pat Robertson would probably intend it as a literal prayer: “Hey, God, how about a little help here?” Knowing something about the two speakers helps figure out what they meant.

If I assume I’m trying to work out what policies you will pursue based on a rather ambiguous statement, I am surely better off incorporating my prior knowledge about your beliefs and history into my interpretation. That’s subject to all sorts of mental biases (partisans can often convince themselves that the wackiest or most offensive statements or actions are somehow innocent when done by their side, and that the most innocent statements are proof of evil intent or stupidity on the other side), but incorporating knowledge about the speaker into your interpretation of their words doesn’t seem irrational at all.

Comments on this entry are closed.

Previous post:

Next post: