Experimentation in Democratic Mechanisms

We examine whether and how democratic procedures can achieve socially desirable public good provision in the presence of deep uncertainty about the benefits of the public good, i.e., when citizens are able to identify the distribution of benefits only if they aggregate their private information. Some members of the society, however, are harmed by socially desirable policies and try to manipulate information aggregation by misrepresenting their private information. We show that information can be aggregated and the socially desirable policy implemented under a new class of democratic mechanisms involving an experimentation group. Those mechanisms reflect the principles of liberal democracy, are prior{free, and involve a differential tax treatment of experimentation group members which motivates them to reveal their private information truthfully. Conversely, we show that standard democratic mechanisms with an arbitrary number of voting rounds but no experimentation do not generally lead to the socially desirable policy. Finally, we demonstrate how experimentation can be designed in such a way that differential tax treatments occur only off the equilibrium path.