Essays in learning and information design

Cabrera, Carlo Antonio (2018) Essays in learning and information design. PhD thesis, London School of Economics and Political Science.

Text - Submitted Version Download (891kB)

Abstract

We re-visit three classical models in information economics. The first chapter studies the screening problem for a seller who owns a single good, and a buyer whose valuation for the good is their private information. We allow for the seller to acquire information at some cost about the buyer’s value, in addition to her choice over the probability of trade and the transfer. The seller thus chooses a Blackwell experiment for each announcement that the buyer makes in a direct revelation mechanism. More informative experiments are more costly. Under mild conditions, there are always optimal mechanisms where the seller acquires coarse information about the buyer. In particular, it is always optimal for the seller to choose an experiment that consists of no more than four signals. When the buyer has only two possible values, the same holds for experiments that consist of at most three signals. The second chapter examines information disclosure in a setting of strategic experimentation. A group of agents continuously and independently choose between a safe arm and risky arms of the same type. When the arms reveal good news, we are able to achieve efficiency in a class of simple information disclosure mechanisms when the agents are initially optimistic enough about their risky arms, but only when there are not too many agents. When the reverse is true, the mechanism must be transparent. Thus, there is a tradeoff between transparency and efficiency. This tradeoff does not exist in the case of bad news. In the final chapter, we borrow insights from social learning theory to understand why institutions have persistent effects. We adapt the classical model in a minimal fashion to accommodate a role for institutions, and demonstrate that social learning is one plausible mechanism of persistence.

Item Type:	Thesis (PhD)
Additional Information:	© 2018 Carlo Antonio Cabrera
Library of Congress subject classification:	H Social Sciences > HB Economic Theory
Sets:	Departments > Economics
Supervisor:	Szentes, Balázs
URI:	http://etheses.lse.ac.uk/id/eprint/3826

Actions (login required)

Record administration - authorised staff only