Collective policymaking panels in the United States often aim to improve accuracy by increasing size and diversity. This study shows that while these elements can enhance group capability, there's a crucial trade-off when they interact.

The research analyzes official revenue forecasts from state independent commissions with consensus group structures. Using regression discontinuity design, it finds an asymmetric effect: adding diversity to large panels reduces forecast accuracy significantly more than reducing diversity in small panels.

This discovery challenges simplistic prescriptions for improving panel performance. It reveals that the relationship between size and diversity is complex:

• Increasing diversity improves accuracy in small groups by about 25%

• But adds error rates nearly fourfold in large groups

Policy implications suggest careful calibration of these characteristics to maximize collective judgment quality.