This article offers the most comprehensive look yet at symbolic racial representation in Congress by analyzing legislators' public speech.

🧾 What speech was examined

• 105,875 constituent newsletters
• 620,838 floor speeches

📊 How speech was compared to constituency change

• Speech referencing Blacks, Hispanics, and Asians was linked to the population shares of those groups in legislators' constituencies.
• Two empirical strategies were used:
• Cross-sectional comparisons relating group size to speech patterns across legislators.
• A difference-in-differences design that exploits redistricting shocks to track within-legislator changes over time.

🔑 Key findings

• Across legislators, White members of both parties are more likely to symbolically represent Black, Hispanic, and Asian groups when those groups are larger in their constituencies.
• These relationships appear in cross-sectional comparisons but do not translate into noticeable within-legislator change: the redistricting-based difference-in-differences tests show little variation in a given legislator's speech as their constituency composition shifts.
• By contrast, substantive representation (legislators' policy behavior) does not vary with group size, indicating a disconnect between symbolic speech and substantive action.

📌 Why it matters

• White legislators are symbolically responsive to constituent racial and ethnic group size in their speech, but this responsiveness is primarily a cross-sectional pattern and does not appear to reflect changing behavior by the same legislators over time or to affect substantive representation.