Political Beliefs and Partisanship

Party Affiliation, Partisanship, and Political Beliefs

• Partisanship is strongly correlated with political attitudes and behavior.
• Researchers conducted a field experiment to investigate the causal effect of party identification.
• A mailing was sent to unaffiliated voters who leaned towards a party.
• The mailing informed recipients that only registered voters could participate in the upcoming presidential primary.
• This treatment group showed a greater tendency to identify with a party and exhibited stronger partisanship.
• Post-treatment, the treatment group demonstrated greater alignment between their latent partisanship and their reported voting behavior and intentions.
• The study supports the idea that partisanship actively influences how citizens behave and perceive the political world.
• Scholars across disciplines, including political scientists, contend that identities like nationalism, religion, and kinship significantly impact attitudes and behavior.
• Partisanship is considered an active force in shaping political evaluations and interactions.
• Studies show a correlation between partisanship and political opinions, assessments of the economy, and interpretations of political events.
• Partisan identities are associated with preferences for biased information and persistence of affiliations over time.
• Social identity theory explains the tendency for humans to distinguish between in-groups and out-groups.
• Membership in a group leads to emotional ties and favoritism for the in-group.
• Identifying with a party is comparable to forming a social identity that shapes attitudes, beliefs, and behaviors with respect to the party's candidates and initiatives.

Research Design, Causal Inference, and the Effect of Partisanship

• Existing research faces challenges in establishing causality between partisanship and political outcomes.
• Cross-sectional data are commonly used but may be prone to omitted variable bias.
• Unobserved factors that correlate with both partisanship and the outcome variable can lead to biased estimation.
• Endogeneity is a concern as political preferences can affect partisanship, not just vice versa.
• Panel data can address unobserved heterogeneity, but assumes unmeasured factors don't impact both variables.
• Panel studies investigate the relationship between change in partisanship and changes in attitudes.
• This approach aims to avoid bias caused by omitted factors, but hinges on assumptions about how unmeasured factors impact variables.
• Experimentation is presented as a solution to inferring causality.
• The 2008 Connecticut presidential primary provided an opportunity for a field experiment.
• A sample of registered but unaffiliated voters were classified as latent Democrats or Republicans.
• The survey data focused on individual political attitudes and behavior.
• A mailing was sent to half the sample reminding them of the need to register with the appropriate party to vote in the primary.
• The experiment measured the impact of induced partisanship on voter attitudes.

Treatment Effect on Party Identification, Party Registration, and Turnout

• Treated latent partisans demonstrated a significantly greater likelihood of identifying with their initially-reported party leanings.
• The treatment increased party registration within that party.
• The treatment led to higher voter turnout in the primary election.
• These effects are statistically significant and appear consistent across latent Democrats and Republicans.

Treatment Effect on Opinions

• The treatment was not uniformly effective across all political attitudes.
• Effects appear stronger regarding voting decisions and evaluations of political figures compared to personal policy views.
• The experiment offers evidence that partisanship influences political outcomes prior to partisan targeting by actors.
• The effects suggest that partisanship influences voting decisions and candidate evaluations, but not necessarily personal policy opinions.

Experimental Protocol

• The experiment was executed in 3 phases
• Identifying latent partisans to serve as the study participants.
• Randomly assigning treatment or control status, with an outreach mailing to the treatment group.
• Conducting a follow-up survey to measure the effects of the partisanship intervention.