Page 37 - 4188

35

nature private information relevant to the relationship. As is typical in
information economics, we refer to the player with the private
information as the informed player informed player and the player
without the private information as the uninformed player.

• Critical to our analysis of these situations is the bargaining game
that determines the contract. We will refer to the contract proposer as the
principal and the player who receives the proposal as the agent. More-

principal over, we assume contracts are proposed on a take-it-or-leave-it
basis: The agent agent’s only choices are to accept or reject the contract
proposed by the principal. Rejection ends the relationship between the
players. A key assumption is that the principal is the uninformed player.

Models like this, in which the uninformed player proposes the contract,
are referred to as screening models. In contrast, were the informed
player the contract screening proposer, we would have a type of

signaling model.
• A contract can be seen as setting the rules of a secondary game to
be played by the principal and the agent.

Given this information structure, the two parties interact according to
some specified rules that constitute the extensive form of a game. In this
two-person game, the players must contract with each other to achieve some

desired outcome. In particular, there is no ability to rely on some
exogenously fixed and anonymous market mechanism. Our focus will be on
instances of the game where the informed player can potentially benefit
from his informational advantage (e.g., perhaps inducing a buyer to pay

more for a good than necessary because she fears the seller is high cost).
But, because the informed player doesn’t have the first move—the
uninformed player gets to propose the contract—this informational

advantage is not absolute: Through her design of the contract, the
uninformed player will seek to offset the informed player’s inherent
advantage.
The Two-Type Screening Model

We will begin to formalize these ideas in as simple a model as
possible, namely the two-type model. In the two-type model, the state of
nature can take one of two-type model two possible values. As is common

in this literature, we will refer to the realized state of nature as the agent’s
type. Given that there are only two possible state, the agent can have one of
just two types.

Before proceeding, however, we need to emphasize that such
simplicity in modeling is not without cost. The two-type model is