Names Signal More Than Race: SES and Migration Bias Audit Experiments

🧭 Research Question and Stakes

Name-based treatments are commonly used to study identity effects—most often race or minority status—on political and social outcomes. These treatments typically assume "information equivalence": that a name signals only the intended characteristic. If names also cue other traits, however, estimated effects cannot be cleanly attributed to the targeted identity.

🔎 How the study was done

• Paired a name perception study with an original correspondence audit experiment targeting U.S. state legislators.
• Measured how different names influence perceptions of multiple characteristics, not just minority status.

📌 Key Findings

• Names systematically alter perceptions of three distinct attributes:
• minority (racial/ethnic) status
• socioeconomic status (SES)
• migrant status
• The correspondence audit shows that perceived low SES lowers reply rates, both across racial categories and within each racial group.
• Because names shift multiple perceptions simultaneously, discrimination observed in audit replies cannot be attributed solely to the intended treatment of minority status.

⚖️ Why It Matters

This evidence demonstrates a violation of information equivalence in name-based treatments: names do more than signal a single intended trait. As a result, estimated treatment effects from name manipulations reflect a multifaceted form of discrimination rather than a pure effect of one characteristic.

🛠️ Recommendation for Researchers

Future studies using name-based or other informational treatments should explicitly account for the possibility that treatments signal multiple characteristics and should reflect that potential violation in research design and interpretation of results.