Computational Methods for the Study of Dynamic Economies

Computational Methods for the Study of Dynamic Economies

Computational Methods for the Study of Dynamic Economies

Computational Methods for the Study of Dynamic Economies

Synopsis

Macroeconomics increasingly uses stochastic dynamic general equilibrium models to understand theoretical and policy issues. Unless very strong assumptions are made, understanding the properties of particular models requires solving the model using a computer. This volume brings together leading contributors in the field who explain in detail how to implement the computational techniques needed to solve dynamic economics models. A broad spread of techniques are covered, and their application in a wide range of subjects discussed. The book provides the basics of a toolkit which researchers and graduate students can use to solve and analyse their own theoretical models.

Excerpt

Thomas J. Sargent

In September 1996 the European University Institute convened the 7th Summer School of the European Economic Association on computational methods for dynamic macroeconomics. Fifty-six graduate students from Europe and 11 faculty from Europe and America gathered for 14 days of lectures and student seminars at the Badia in Fiesole. This book assembles papers whose first drafts served as the lecture notes for the conference. The conference aimed to disseminate ideas about how to compute solutions of dynamic economic models and how to apply them. Faculty and students took turns talking and listening. Algorithms were discussed and computer programs supplied. Teachers gave two lectures a day. The remainder of each day was filled with student presentations of their work. More work was done informally over lunch and coffee tables. The papers in this volume capture some of what we, the teachers, told the students. Unfortunately they cannot reflect the energy, interest, and promise that the students showed us in their questions and their presentations of their accomplishments and their plans. The conference and this volume are about ways to circumvent or confront the 'curse of dimensionality' associated with the functional equations that solve dynamic equilibrium models. Though they are designed to find rational expectations equilibria, before they have converged, computational algorithms resemble learning algorithms. Successive iterates of the algorithm are approximations that adapt to information about approximation errors. Several of the teachers at the conference (Uhlig, Marcet, McGrattan, Marimon, and I) have used what we know about computational algorithms to think about adaptive models of bounded rationality. Marimon lectured on some of this work at the conference, and other lecturers touched on it. The same tools that support computation of rational expectations models will be relied upon to venture beyond rational expectations, when we choose to do so. An air of enthusiasm about computational dynamic economics came to the conference from the beautiful setting overlooking Florence, the presence and pleasant interruptions of the youngest conference participant Zoe McCandler in her stroller, and the commitment of our students.

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.