Shared Demographics Don’t Boost Persuasion in Political Conversations

What the Authors Ask

David Broockman, Joshua Kalla, Nicholas Ottone, Erik Santoro, and Amanda Weiss test a common assumption in political persuasion: are people more likely to change their minds when the person talking to them shares their demographic characteristics (age, gender, or race)? Practitioners and scholars often expect shared identity to increase trust and hence persuasiveness, so this question matters for the design and scaling of one-on-one conversation interventions.

Why It Matters

Many civic programs and campaigns rely on interpersonal conversations to influence voters. If demographic matching reliably increases persuasion, implementers would need to recruit and match messengers by age, gender, or race—raising costs and complicating scale-up. Conversely, if shared demographics do not matter, conversations could be deployed more flexibly across diverse populations.

How the Study Was Done

• The authors synthesize evidence from eight randomized experiments involving interpersonal conversations on four distinct political topics, totaling 6,139 participants.
• Experiments varied whether the speaker and recipient shared demographic traits (age, gender, race) and measured subsequent attitude change.
• The paper also reports a survey of practitioners and scholars documenting common expectations that shared demographics should aid persuasion.

What They Found

• Contrary to common expectations, shared demographics (age, gender, or race) do not meaningfully increase the persuasive effects of interpersonal conversations in these experiments.
• The results align with dual-process theories suggesting messenger attributes operate as peripheral cues that matter less when listeners engage in effortful, conversational processing.

What This Means for Practice and Research

These findings suggest that conversation-based interventions can be effective without tight demographic matching of messengers and recipients, which simplifies program design and may improve scalability. For scholars, the results challenge assumptions about identity-based messenger effects and suggest further work on when and why messenger characteristics matter across political contexts and topics.