This paper proposes a model of ambiguous language. We consider a simple cheap talk game in which a sender who faces an ambiguity averse receiver is able to perform ambiguous randomization, i.e. to randomize according to unknown probabilities. We show that for any standard influential communication equilibrium there exists an equilibrium featuring an ambiguous communication strategy which Pareto-dominates it in terms of consistent planning ex ante utilities. Ambiguity, by triggering worst-case decision-making by the receiver, shifts the latter's response to information towards the sender's ideal action, thus encouraging more information transmission

We consider three mechanisms for the aggregation of information in heterogeneous committees voting by Unanimity rule: Private Voting and voting preceded by either Plenary or Subgroup Deliberation. While the first deliberation protocol imposes public communication, the second restricts communication to homogeneous subgroups. We find that both protocols allow to Pareto improve on outcomes achieved under private voting. Furthermore, we find that when focusing on simple equilibria under Plenary Deliberation, Subgroup Deliberation Pareto improves on outcomes achieved under Plenary Deliberation

We study cheap talk communication in a simple two actions-two states model featuring ambiguous priors. First, we find that in equilibrium, S often mixes between messages triggering different beliefs and behavior by R while R also occasionally randomizes. We interpret randomization by respectively S and R as embodying two different modes of ambiguous communication. Second, we find that for sufficiently high ambiguity, more than two messages are often necessary to implement the optimal decision rule of S. If only two messages are available and S faces his preference twin, S may be unable to implement his optimal decision rule and influential communication may be altogether impossible. We stress the non replicability of these results within an expected utility environment. Third, we show that the addition of a little ambiguity may generate influential communication that is unambiguously advantageous to S

We generalize the classical binary Condorcet jury model by introducing a richer state and signal space, thereby generating a concern for consistency in the evaluation of aggregate information. We analyze truthtelling incentives in simultaneous pre-vote communication in heterogeneous committees and nd that full pooling of information followed by sincere voting is compatible with a positive probability of ex post con ict in the committee.

We study experimentally the effectiveness of communication in common value committees exhibiting publicly known heterogeneous preferences and test whether social preferences or cognitive constraints drive the (non-)existence of strategic communication. As prior communication may affect voting decisions, we separately and jointly test communication and voting choices and how they depend on the presence of heterogeneous preferences. Results are only consistent with a model of cognitive heterogeneity. Roughly 80% of subjects truth-tell and use a decision heuristic (i.e. vote with the majority of announced signals). The remaining sophisticated agents lie strategically and approximately apply their optimal decision rule.