Are Two Heads Better Than One? Monetary Policy by Committee

TLDR

Group decision‑making theories that predict a group follows its average, median, or best member fail to explain the observed outcomes. The study aimed to test whether groups take longer to decide than individuals. One experiment replicated real‑life monetary policy decision processes. Both experiments found that groups are not slower than individuals and that group decisions are on average better than individual decisions, regardless of unanimity or majority rule, suggesting that group interactions enhance decision quality.

Abstract

Two experiments were conducted to test the common hypothesis that groups make decisions more slowly than individuals. One of these experiments imitates real-life monetary policy decisions. In both cases, the hypothesis is found wanting: groups are not slower than individuals. In both experiments, we also find that group decisions are on average better than individual decisions. This holds regardless of whether the groups make decisions by unanimity or majority rule. Simple mechanical theories of group decisionmaking— that the group follows its average player, median player, or best player—do not explain the results. Group interactions seem to matter.