Screening and the Welfare Trade-offs of Privacy

Abstract

A principal and agent interact in a class of models a la Mussa and Rosen (1978) with
continuously divisible product quality. Prior to their interaction, a notional planner
designs the information disclosure environment which determines the extent to which
the principal can observe the willingness-to-pay of the agent. We contrast the class
of Pareto optimal information disclosure policies when the willingness-to-pay of the
consumer is exogenous and when, instead, the distribution of agent types is determined
by an unobservable and costly investment by the agent. In either class, full
privacy may or may not be Pareto optimal, and within the class of Pareto optima,
there is always a strict tradeoff between agent welfare and total welfare (as measured
by agent surplus plus principal surplus). In a two-type model, this tradeoff is fully
parameterized by a one-dimensional measure of information disclosure. Full disclosure
is always Pareto optimal with exogenous willingness-to-pay and never Pareto
optimal with endogenous willingness-to-pay.

Keywords: Screening, Privacy, Price discrimination,Welfare

JEL Classification: D61, D82, D83, D86, L12