🚨New working paper w/ some terrific Stony Brook grad students!
Using a pre-registered experiment, we test whether shifts in groups' voting behavior can change partisans' sentiments toward members of those groups (Black, LGT & Muslims).
Answer? No.
polisky
preprints.apsanet.org/engage/apsa/...
Thank you! 🙏
Yes, agreed. Though the other interesting piece is that feelings also don’t get more negative when the group moves *away* from one’s party. For example, Dems don’t show more negativity toward LGT even if LGT start voting more Republican.
I understand why a 5% shift is large in practical terms in close elections, but it doesn't shift the overall partisanship of the target group (I think?).
I guess, I wonder if people distinguish between groups voting at 89% vs. 84% for Ds? Is a group perceived much less D in these conditions?
Great Q, Mark (fan of your work!). Thank you!
I suspect this will be a common reaction. FWIW, we tried hard to balance a test of the hypothesis w/ something actually plausible. It would be hard to know what to make of sig effects based on implausible shifts, or if they've even be trusted by Ss.
We note in the Discussion that (1) larger (but less plausible) shifts, and/or (2) lesser known groups, could both find sig effects. But for these very prominent groups in U.S. politics, using shifts that (in reality) would be enormous, we don't find any effects.
Also FWIW, we presented and framed these shifts as being "huge", AND a manipulation check found that respondents did indeed view these to be pretty large shifts.
But still, it's an open question whether, say, a group's change of 40% and/or being majority Republican, might find an effect.
That all makes sense. It’s an interesting study and definitely informative. I don’t have a better suggestion for balancing shift size and believability.
Just thinking in terms of hypotheses about group coalitions and stereotypes and whether that change moves the needle on those
Yep, definitely all fair points. I believe it could work, but (based on these results) might need to venture into the realm of less plausibility and/or relevance.
More narrowly, I think it also speaks to whether media reports of groups' shifts (eg, Blacks for Trump) can change attitudes (prob not)
