Abstract
This dissertation consists of three essays. Essay 1: Lobbying, Information, and Private and Social Welfare. Lobbying is potentially welfare improving if it is a form of information transmission. This essay studies when and for whom this is the case. In the model, a government (G) can redistribute income between two interest groups (A and B). Only one of them (A) has the opportunity to lobby. Lobbying is understood as A’s acquiring and strategically revealing policy-relevant information to G. Among the results is that A may be worse off as a consequence of lobbying while B is better off. It can also be the case that lobbying makes both the active and passive interest groups and the government all worse off. Essay 2: Are we better off if our politicians know how the economy works? This essay concerns public policy and welfare in a society where citizens’ preferences over public policy depend, in varying degrees, on some unknown state of the world. That is, people are heterogeneous with respect to their responsiveness to the unknown state. Public policy is decided on by a policymaker who is elected among the citizens by majority vote. Given this framework it is asked whether the citizens would be better off if the amount of uncertainty that the policymaker is facing were smaller. Among the results is that those who are sufficiently responsive to the unknown state may be worse off if the variance of the stochastic variable decreases. Essay 3: Incomplete information in the Samaritan’s dilemma: The dilemma (almost) vanishes. Suppose that an altruistic person, A, is willing to transfer resources to a second person, B, if B comes upon hard times. If B anticipates that A will act in this manner, B will save "too little" from both agents’ point of view. This is the Samaritan’s dilemma. The mechanism in the dilemma has been employed in an extensive literature, addressing a wide range of both normative and positive issues. However, this essay shows that the undersaving result is not robust to the assumption that information is complete: by adding an arbitrarily small amount of uncertainty one can sustain an equilibrium outcome that is arbitrarily close to ex post incentive efficiency. One may also sustain outcomes with oversaving.