Dissertation/Thesis Abstract

Policy Design for Competitive Retail Electric Institutions: Artificial Intelligence Representations for a Common Property Resource Approach
by Pandit, Nitin S., Ph.D., George Mason University, 1998, 161; 10242887
Abstract (Summary)

The U.S. electricity industry is being restructured to increase competition. Although existing policies may lead to efficient wholesale institutions, designing policies for the retail level is more complex because of intricate interactions between individuals and quasi-monopolistic institutions. It is argued that Hirshman's ideas of "exit" and "voice" (Hirshman, 1970) provide powerful abstractions for design of retail institutions. While competition is a known mechanism of "exit," a novel design of the "voice" mechanism is demonstrated through an artificial intelligence (AI) based software process model. The process model of "voice" in retail institutions is designed within the economic context of electricity distribution — a common property resource (CPR), characterized by technological uncertainty and path-dependency. First, it is argued that participant feedback (voice) has to be used effectively to manage the CPR. Further, it is noted that the decision process, of using participant feedback (voice) to incrementally manage uncertainty and path-dependencies, is non-monotonic because it requires the decision makers to often retract previously made assumptions and decisions. An AI based process model of "voice" is developed using an assumption-based truth maintenance system. The model can emulate the non-monotonic decision making process and therefore assist in decision support. Such a systematic framework is flexible, consistent, and easily reorganized as assumptions change. It can provide an effective, formal "voice" mechanism to the retail customers and improve institutional performance.

Indexing (document details)
Advisor: Haynes, Kingsley E.
Commitee: Button, Kenneth J., Fuller, Stephen S., Gulledge, Thomas R., Stough, Roger R.
School: George Mason University
Department: Public Policy
School Location: United States -- Virginia
Source: DAI-A 78/06(E), Dissertation Abstracts International
Subjects: Public policy, Energy, Artificial intelligence
Keywords: Common property resource, Decentralization, Policy design, Retail electric institutions, Truth maintenance
Publication Number: 10242887
ISBN: 9781369533903
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy