44 Pages Posted: 25 Feb 2014 Last revised: 24 Apr 2014
Date Written: February 24, 2014
This Article identifies and examines two alternative institutional structures for hierarchical institutions — “bounded” vs. “unbounded” institutional structures. In a bounded structure, a principal decides on a bounded numerical allocation and then an agent allocates to subjects while complying with the bound. In an unbounded structure, the principal provides no numerical cap or floor on agents, but instead provides some guidance to the agents regarding allocation. An example of a bounded institution is grading to a pre-arranged curve, while an example of an unbounded institution is granting a particular grade to every student who meets a particular threshold.
Bounded and unbounded institutions differ in their strengths and weaknesses. Principals should choose bounded institutions when there are consistent and large populations, when agents are likely to make systematic errors but otherwise share rank order preferences with the principal, and when it is difficult to devise rules along other dimensions of the agent’s decision. If agents are biased but share a ranking order with the principal and principals know population traits but not individual traits, bounded institutions can produce a perfect allocation even though neither the principal nor the agent is fully informed or free of error. In other words, with many students and shared quality standards, grading to a curve can produce ideal grading, even if some professors are inclined to be generous with their grading while others are stingy.
Suggested Citation: Suggested Citation