Three essays on the political economy of information.

Summary This paper develops three formal models of collective choice, and in particular tries to show that it can be useful to consider political conflicts as not only conflicts of interests but also conflicts of beliefs. The first model attempts to account for an empirical regularity that has been rigorously established but never explained theorically: the tendency for political conflicts to be one-dimensional. It considers a populated world of agents characterized by identical preferences but different initial beliefs about the effects of different possible collective choices, and analyzes the strategic communication of information about this uncertainty. The repetition of this communication allows to consider the distribution of beliefs as a stochastic process whose asymptotic properties are studied. The main result establishes the convergence of the beliefs to a unidimensional distribution, whatever the geometry of the initial conflict. The second model studies the transmission of information about a unidimensional collective decision in a world where individuals have divergent interests. It shows that information transmission tends to be perfect when preferences tend to coincide. On the contrary, it tends to become very imprecise when the population grows, to the point that agents use only two messages. The third model deals with the political economy of federalism. It mainly studies the extent of redistribution in a federation as a function of the type of federal institution - federal democracy or intergovernmental bargaining - and shows that the extension of available fiscal tools can make it more difficult to form a federation if it facilitates the expropriation of one nation by the others.