Deliberation and Polarization: Evidence from Group Decision Experiments

How does deliberation in like-minded groups impact collective decision-making - does it polarize or moderate voting decisions? To investigate this question, we developed a pre-registered laboratory experiment (N = 216) using a three-round group allocation design that systematically varies group composition (like-minded vs. mixed-minded) and decision rules (unanimity vs. median). Participants allocated real monetary endowments between themselves and a local refugee organization, enabling us to study morally salient decisions regarding redistribution. Contrary to enclave-deliberation theories, we find that deliberation consistently moderates allocations toward fairness benchmarks, regardless of group composition or rule. Selfish participants in mixed-minded groups, in particular, reduced self-retention most strongly, while generous participants slightly increased it. These moderation effects vanish once participants return to individual decision-making. The results reveal deliberation as a temporary depolarizing mechanism that promotes short-term fairness and consensus without lasting attitudinal change. This finding adds to research on behavioral models of deliberation by highlighting the social-norm dynamics that underlie moderation in collective decisions.