The Effects of the US Presidential Election on Affective Polarization: Pre-Analysis Plan

How to cite

Höhne, Christian; Gallo, Edoardo; Aidt, Toke. 2024. "The Effects of the US Presidential Election on Affective Polarization: Pre-Analysis Plan." AEA RCT Registry. AEARCTR-0014156. https://doi.org/10.1257/rct.14156-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Christian Höhne* , Edoardo Gallo , Toke Aidt

* Primary investigator

Registry ID
AEARCTR-0014156
DOI
https://doi.org/10.1257/rct.14156-1.0
First registered
Last updated
Status
In development
Trial dates
2024-08-13 – 2025-02-28
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/14156

Abstract

The purpose of this research project is twofold. We first aim to compare incentivized measures of polarization – captured as behavior in trust & prisoner dilemma games – against commonly used survey-elicited measurements of polarization. To do so, we employ an event-study approach, comparing polarization levels at an early stage of the presidential campaign and around the US presidential election. This also enables us to address our second research goal, namely to experimentally investigate the effect of the US presidential election process on affective polarization.

Experimental design

Through our experimental set-up, we can elicit some of the key underlying factors of affective polarization. The Trust Game (TG) allows us to elicit participants’ levels of trust and reciprocity, while in the Prisoner Dilemma Game (PDG), we can see if participants choose to cooperate with their partners. By informing participants about the partisanship of the participant they are playing with, we can test the effect of partisan identity on behavior and thereby elicit an experimental measure of affective polarization.
This experimental measure can then be benchmarked against commonly used survey measures of affective polarisation, such as the ANES thermometer question, social distance and trust questions.

Design details

Randomization method
We randomize the order in which participants play the games per computer.
Randomization unit
There are two levels for our treatments: 1. We randomise the order in which participants play the games. This is done on the individual level. 2. We also use participants' state of residence as identifying variation in our analysis.
Clusters
10 states
Observations
1,200 observations overall. 400 observations per experimental wave.
Arms
5 treatment states and 5 control states 600 observations per treatment arm and 200 observation per treatment arm and experimental wave.
Primary outcomes
We are looking at affective polarisation measured by survey responses and experimentally incentivized behavior: - ANES thermometer measure - Survey Trust measure - Social Distance measure - Experimental Trust measure - Experimental Trustworthiness measure - Experimental Cooperation measure

Cited by

No citation data available for this registration yet.

Keywords