Political economy in imperfect information: three essays on planning and indirect democracy.

Summary In this paper, we study three specific problems in normative public economics or positive political economy related to the lack of information (or the impossibility of obtaining information) about the state of the economy. First, in the classical optimal income taxation model of Mirrlees, we abandon the assumption that the state knows perfectly the distribution of talents in the economy. In a dynamic model with Bayesian learning, we show that this initial uncertainty leads to a systematic bias towards a higher inactivity rate (chapter 2). Second, we continue to explore the consequences of imperfect information about talent allocation by considering the case where the decision is not taken by a central planner but by democratically elected representatives based on their fiscal policy proposals. We show that when two political parties compete on the basis of proposals for a linear income tax, the lack of certainty about the distribution of talent leads to higher inactivity and tax rates than when this distribution is known with certainty at the outset (chapter 3). In a final chapter, we address the problem of the informational efficiency of indirect democracy: in a framework close to that of the Condorcet jury, we show that even if voters are rather poorly informed about the economic situation, the competition for power encourages relatively better informed political parties to propose the policy they consider best for the electorate (chapter 4).