Rarely, if ever, do parties contemplating a joint project commit resources without engaging in nonbinding discussions about who does what. To examine the role of such discussions, we model the investment decision as a voluntary contribution game where each player is privately informed of her benefit from project's completion. Efficiency of every equilibrium of this game is improved when a prior stage of communication is allowed. Interestingly, this improvement can be achieved in "simple equilibria" where a player simply announces whether (not how much) she plans to contribute.With symmetric priors, the simple equilibrium that maximizes either utilitarian ex ante welfare or the probability of completion has a curiously egalitarian feature-whenever the project is completed, each party contributes exactly half the cost, independent of private information.

A partially informed expert, A, strategically transmits information to a principal, P. The residual uncertainty faced by the expert effectively causes the bias between P and A to be random, with two consequences. First, by misreporting, A is likely to induce a decision choice by P, after the resolution of the residual uncertainty, that is either close to A's ideal position or too far from it, whereas truthful reporting keeps such variations "more balanced". A convex loss function of A thus favors truthful reporting. Second, by retaining authority of decision making and communicating with A, P avoids exposure to risks due to residual uncertainty.Better information transmission and the associated insurance benefit thus often imply P preferring control over delegation, despite A having superior information.JEL Classification: C72, D82, D83.Key Words: Authority, Delegation, Effective Bias, Full Revelation, Partially Informed Expert, Strategic Information Transmission, Multidimensional Uncertainty. * We thank Bob Anderson, Ricardo Alonso, Bob Gibbons, Atsushi Kajii and Navin Kartik for their very helpful comments and conversations. This work has also benefited from presentations at various seminars and conferences.

We study the dual issues of allocation and coalition formation in a model of social learning. For a class of economies which can be expressed in terms of a real valued characteristic function, we first show that all self-perpetuating allocations realized from a simple bargaining game must be core allocations although players make simultaneous demands for surplus and only on their own behalf. Following this, we provide a sufficient condition under which the society eventually learns to divide the surplus according to some core allocation, regardless of the initial history.

No abstract

No abstract