Abstract: How does jury deliberation affect the pre-deliberation judgments of individual jurors? Do deliberating juries reduce or eliminate the erratic and unpredictable punitive damage awards that have been observed with individual jurors? In this paper we make progress on these two questions, in part by reporting the results of a study of over 500 mock juries composed of over 3000 jury eligible citizens. Our principal finding is that juries did not produce less erratic and more predictable awards than individuals, but actually made the problem worse, by making large awards much larger and small awards smaller still, even for the same case. Thus, a key effect of deliberation is often to polarize individual judgments, a pattern that has been found in many other group decision making contexts. This finding of polarization--the first of its kind in the particular context of punitive damage awards--has important implications for jury awards involving both punitive and compensatory damages, and raises questions about the common belief that groups, and in particular juries, generally make better decisions than individuals.