Positive Feedback and Path Dependence Using the Law of Large Numbers

Abstract

Economists have become interested in the behavior of random processes with positive feedback but have sometimes found it difficult to introduce students to this research. Simulation of the law of large numbers with increasing amounts of feedback provides a convenient framework for such discussion and facilitates a nontechnical introduction to recent work on path dependence.

Key Words:

• law of large numbers
• path dependence
• positive feedback
• simulation

Additional information

Notes on contributors

Peter Hans Matthews

Peter Hans Matthews (e-mail: [email protected]) is an assistant professor of economics at Middlebury College. The first draft of this article was included in “Simulation and the Teaching of Time Series Econometrics,” a paper co-authored with John Elder and presented at Teaching Undergraduate Econometrics: Does Monte Carlo Work? a conference held at Middlebury College in November 1999. The author completed this revision as a Visiting Scholar at UC San Diego, with generous leave support from Middlebury College. He thanks Neviana Petkova, a Christian A. Johnson summer intern, for research assistance, as well as William Becker, David Colander, Carolyn Craven, John Elder, William Greene, Peter Kennedy, Michael Murray, Michael Olinik, and three anonymous reviewers for their comments.