Quantifying the potential persuasive returns to political microtargeting

Area of Science:

• Political Science
• Computational Social Science
• Communication Studies

Background:

• Concerns exist regarding political microtargeting's influence on elections and democracy.
• Limited empirical research quantifies microtargeting's persuasive advantage over alternative campaign methods.

Purpose of the Study:

• To estimate the persuasive advantage of microtargeting compared to other campaign strategies using U.S. policy issue advertising.
• To investigate the role of machine learning and message pretesting in optimizing microtargeting effectiveness.

Main Methods:

• Two survey experiments were conducted to compare microtargeting with two alternative messaging strategies.
• Microtargeting involved combining machine learning with message pretesting to tailor ads to individuals.
• Performance was evaluated based on persuasive impact on policy attitudes.

Main Results:

• The microtargeting strategy demonstrated an average persuasive advantage of 70% or more over alternative strategies for a specific policy attitude (Study 1).
• No additional persuasive gains were observed when targeting messages by more than one covariate.
• The performance advantage of microtargeting was issue-specific and more limited when used to identify target attitudes (Study 2).

Conclusions:

• Microtargeting, integrating message pretesting and machine learning, can enhance campaign persuasive influence.
• Effective microtargeting may not necessitate extensive personal data collection to identify complex audience-message interactions.
• The actual persuasive advantage of microtargeting is context-dependent and varies across different political issues and targeting approaches.