56,032 rated or chosen profiles from 3,502 respondents, in up to 8 tasks of 2 profiles varying on 5 attributes.

About this table

Description	US adults (n = 3,502) chose, or abstained, between pairs of hypothetical candidates described by occupation, office experience and three campaign statements (populist or liberal-democratic rhetoric on pluralism and moralism, and a policy position); 8 pairs each.
Reference	Kustov, A., & Dai, Y. (2026). What is populism good for? An experimental test of mobilization effects. Research & Politics.
DOI	10.1177/20531680261434109
Licence	CC0 1.0
Source data	https://doi.org/10.7910/DVN/3OMMOS

Notes

Conjoint table (irw_conjoint_2). choice = candidate chosen in the pair; pref = 0 (abstain, neither candidate) is the opt-out, 4,516 of 28,016 tasks; 8 tasks of 2 candidates, 5 attributes (job, office, and three statements: pluralism, moralism, policy, each drawn from a bank of 32/32/64 sentences). The pluralism statement is blank on 3.0% of candidates, only with non-populist moralism, and is coded (not shown) as an inference (the deposit is silent). The vote question wording is not in the deposit; respondent count (3,502) matches the authors' replication notebook.

These notes describe the source data, which the IRW reproduces as released. All notes are listed on the Data Notes page.

Size and shape

Respondents	3,502
Rows (one per respondent, task and profile)	56,032
Tasks per respondent (most)	8
Profiles per task	2
Attributes	5
Tasks with no profile chosen	4,516

Design

Topic	candidate choice
A profile represents	person
Country	US
Language shown	en
Language of stored levels	en
Randomization restrictions	observed in the data (not documented by the source): the pluralism statement is absent only when the moralism statement is a non-populist one (1,703 of 28,128 such candidates); populist moralism and populist pluralism statements always appear together
Task order	recorded in the deposit
Profile position	recorded in the deposit

Because the attribute levels were not randomized independently, an estimator has to account for the restriction; a plain difference in means across levels can mislead.

Outcomes:

choice	Which candidate would you vote for in this pair (or abstain)? (paraphrase; rep.Rmd: voted = voted for either candidate, 0 = abstained) [choice, opt-out allowed]

Columns

attr_jobattr_moralism_statementattr_officeattr_pluralism_statementattr_policy_statementchoicecov_collegecov_gendercov_ideologycov_party_ididprofiletask

The layout (one row per respondent, task and profile; choice/rating outcomes; attr_ columns holding each attribute level as displayed) is defined in the IRW conjoint standard. What the outcomes asked is under Design above.

Get the data

Download CSVno account needed Browse on Redivisexplore and query Croissant metadataHugging Face, Kaggle, OpenML

Or load it directly in R or Python (both need a free Redivis account; the first call opens a browser to sign in):

# R
install.packages("remotes")
remotes::install_github("itemresponsewarehouse/Rpkg")
library(irw)
df <- irw_fetch("kustov_2026_populism_mobilization", source = "conj")

# Python
pip install irw

import irw
df = irw.fetch("kustov_2026_populism_mobilization", source="conj")

How to cite

@article{Kustov_2026, title={What is populism good for? An experimental test of mobilization effects}, volume={13}, ISSN={2053-1680}, url={http://dx.doi.org/10.1177/20531680261434109}, DOI={10.1177/20531680261434109}, number={1}, journal={Research &amp; Politics}, publisher={SAGE Publications}, author={Kustov, Alexander and Dai, Yaoyao}, year={2026}, month=Jan }

Version and provenance

IRW version	v584
Redivis dataset	irw_conjoint_2 v1.4
Redivis dataset DOI	10.57761/hk97-wv30
Manifest pin for this IRW version	v1.4
Metadata source	irw_meta v44.0