Crooked Hillary and Sleepy Joe: name-calling’s backfire effect on candidate evaluations

ABSTRACT

Throughout his political career Donald Trump has utilized name-calling when referring to his opponents. These pejoratives are a ubiquitous part of political discourse in contemporary society. Scholarly research has yet to examine the effect that this type of incivility has on individuals’ evaluations of both the attacker (i.e. the person using name-calling) and the victim. Our research aims to fill this gap by testing the effect of name-calling through the implementation of a national survey experiment. We test the effect of name-calling on candidate evaluations by randomly inserting a pejorative in front of a fictitious candidate’s name in a news story. Our findings indicate that name-calling often backfires. Respondents who saw the pejorative tend to rate the attacker lower. Our findings also show an odd partisan symmetry in how respondents rate this behavior by their co-partisans, i.e. both Republicans and Democrats punish Democratic candidates that use name-calling but ignore Republicans’ use of it.

Disclosure statement

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

Notes

1 The notable exception is during his presidential debate performances, which are aimed at the viewing audience.

2 All survey material were reviewed by the XX Institutional Review Board (protocol # 1905759111) and given “exempt” status.

3 We contracted with the survey research organization Dynata (formerly known as Survey Sampling International) to provide respondents to the survey we designed and hosted using Qualtrics' survey platform. Dynata maintains panels of respondents intended to be representative of the U.S. population. The survey was in the field from July 13, 2019 to July 24, 2019.

4 We only find minor differences in how the two pejoratives perform in our experiment.

5 As expected, the intraclass correlation coefficient on a null model with “score” at level 1 and “subject id” at level 2 is large (0.318), indicating there is substantial clustering in our data.