Beyond Residency: Variations in Local Ties and Electoral Preferences in Mayoral Elections

Last registered on July 06, 2026

View Trial History

Pre-Trial

Trial Information

General Information

Title

Beyond Residency: Variations in Local Ties and Electoral Preferences in Mayoral Elections

RCT ID

AEARCTR-0019103

Initial registration date

July 06, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

July 06, 2026, 9:41 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Germany

Region

Primary Investigator

Name

Sebastian Schirner

Affiliation

ifo Institute, Munich

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Klara Lehmann

PI Affiliation

ifo Institute, Dresden branch

Contact Investigator

Additional Trial Information

Status

In development

Start date

2026-07-07

End date

2026-07-21

Keywords

Behavior, Electoral

Additional Keywords

Local Political Economy

JEL code(s)

D72;H70;H72;D91

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

This research project examines local ties and the electoral success of mayoral candidates in local elections. We employ a conjoint survey experiment on electoral candidate preferences among voting-age residents (18-75 years) in Germany. Fielded by a German survey institute, the survey aims to quantify voters’ relative preferences for mayoral candidate attributes, with a particular focus on locality markers in local elections. We conceptualize local identity as a multidimensional construct encompassing ascribed (birthplace), achieved (length of residence), and performative (local civic engagement) dimensions, which are experimentally varied in the conjoint design. The study examines how these dimensions shape perceptions of candidates as members of the local in-group and, more broadly, how local identity structures group boundaries in electoral choice.

External Link(s)

Registration Citation

Citation

Lehmann, Klara and Sebastian Schirner. 2026. "Beyond Residency: Variations in Local Ties and Electoral Preferences in Mayoral Elections." AEA RCT Registry. July 06. https://doi.org/10.1257/rct.19103-1.0

Sponsors & Partners

Interventions

Intervention(s)

The study employs a forced-choice paired-profile conjoint experiment embedded in a representative online survey. Respondents repeatedly choose between two hypothetical mayoral candidates. In addition to the forced-choice decision in each conjoint task, respondents provide a rating-based evaluation on a 7-point scale indicating how likely they would vote for the respective candidate profile as mayor. This rating-based outcome is collected after each task for each candidate profile and serves as a secondary outcome that complements the binary choice measure. This paired-profile design closely mirrors real-world local elections, in which—despite the formal presence of multiple candidates—electoral competition often concentrates on two viable contenders, particularly in runoff settings.

Intervention (Hidden)

Intervention Start Date

2026-07-07

Intervention End Date

2026-07-21

Primary Outcomes

Primary Outcomes (end points)

Candidate choice (binary variable)

Primary Outcomes (explanation)

Respondents are forced to choose between the two hypothetical candidate profiles. The outcome is the resulting binary variable.

Secondary Outcomes

Secondary Outcomes (end points)

Candidate ratings (7-point Likert scale)

Secondary Outcomes (explanation)

Respondents rate both hypothetical candidate profiles on a 7-point Likert scale.

Experimental Design

The study employs a forced-choice paired-profile conjoint experiment. Respondents choose between and rate hypothetical profiles of mayoral candidates. The profiles contain eight attributes with randomized attribute levels. Attributes have at minimum two and at maximum five levels. Attribute order is randomized across respondents but constant within respondents. Respondents complete seven conjoint tasks.

Experimental Design Details

Randomization Method

Attribute levels and attribute order across respondents are uniformly randomized subject to plausibility constraints. Randomization is built into the survey programming.

Randomization Unit

Randomization happens at the candidate profile level.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

2,500 respondents

Sample size: planned number of observations

2,500 respondents x 2 profiles x 7 tasks = 35,000 observations

Sample size (or number of clusters) by treatment arms

2,500 respondents

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

We assess statistical power using two-sided tests with α = 0.05 and the conventional 80% power target. With R = 2,500 respondents, J = 7 tasks, and K = 2 profiles, the raw number of profile evaluations is nraw = R x J x K = 35,000. To account for within-respondent clustering, we apply a design-effect correction with cluster size m = J x K = 14 and an intra-cluster correlation of ρ = 0.05, giving DE = 1 + (m − 1)ρ = 1.65 and an effective sample size of neff = nraw/DE ≈ 21,212. The minimum detectable effect at 80% power is MDE = (z1−α/2 + z1−β ) SE. For the eight conjoint attributes, the minimum detectable AMCE ranges from approximately 1.9 percentage points (two-level attributes) to 3.0 percentage points (five-level attributes such as residence). These MDEs lie below the 3–8 pp AMCEs typically reported in the candidate-locality literature, so the main-effect tests (H1–H3) are well powered.

Supporting Documents and Materials