Incentives and Organizations in Markets

Each class is divided into two blocks, A and B. The A block consists of a survey of a key branch of the literature on incentives and organizations and markets. Block B consists of a detailed look at one of the instructors contributions to this literature. The goal of the class is to transition students from reading about research to producing it. Each of the A blocks is designed to familiarize you with a set of models that are widely used in applied settings. You should think of these as "Lego blocks" for constructing research papers. They are basic models that you can build upon in your own original research, both empirically and in applied theory. The B block is meant to show how such tools may be applied in novel ways to produce new insights.

A student or group of students will present the A block for each class. This presentation, lasting 80 minutes, accounts for 25% of the class grade. It should provide motivation, highlights of the literature, layout the general class or classes of models, prove main results, and offer some open questions or suggestions for future research direction, i.e. it should be at the level of an author of the survey presenting the work in seminar.

Following the conclusion of the A block, there will be a break of 20 minutes for coffee and informal discussion. After this, the instructor will present in the B block. The format will be similar: an 80 minute presentation motivating the particular problem, connecting it to the literature, working out the model, and presenting the results.

Since the format of both blocks is in seminar style, participants are encouraged to ask questions. Participation in this aspect of the class counts for an additional 25% of the grade.

Finally, since the goal of the class is to transition students from consuming research to producing it, students must write a short research paper based on the concepts learned in class. The research paper determines the remaining 50% of the course grade. This paper may be theoretical or empirical. If empirical, it must use real or simulated data and contain actual empirical results.

The concept of simulated data may be unfamiliar to some. For the purposes of the course, an acceptable empirical paper is one that identifies a data source, even if the data from this source are not yet available. The paper then hypothesizes some noisy data generating process underlying this data. This could be based on broad, structural features of the setting or agent-based simulations. The student will then program this data generating process into a favorite program (anything from Excel to C++ is acceptable) and generate a simulated data set. All empirical analyses will then be performed on the simulated data.

Updates, links to certain course materials, and other items will be posted on the course webpage.

The syllabus groups items by (approximately) weekly topic.

Structure of the Course

Each week, a student or group of students will present based on the readings. Following this, I will present an overview of the types of research to come out of this class of models as well as highlight any open questions. The goal of the class is to transition students from reading about research to producing it. Each of the units is designed to familiarize you with a set of models that are widely used in applied settings. You should think of these as "Lego blocks" for constructing research papers. They are basic models that you can build upon in your own original research.

Resources for Agent-Based Simulation:

Leigh Tesfatsion's home page has a wealth of material, and the following is a reasonable place for a newcomer to agent-based simulation to start. Check out her online guide to agent-based modeling.

She writes: The purpose of this on-line guide is to suggest a short list of introductory readings and supporting materials to help newcomers become acquainted with Agent-Based Modeling (ABM). Our primary intended audience is graduate students and advanced undergraduate students in the social sciences.

Main ideas: The revelation principle forms the cornerstone for the theory of contracts and incentives in organizations. This topic reviews the revelation principle, the principal families of contract theory models, and the mechanism design paradigm.Nearly all forms of interaction among competing agents may be modeled as an auction of some form. The starting point for this work is the so-called independent private value (IPV) model of auctions along with its main result, the Revenue Equivalence Theorem.

Chapters 8.E and 13 of Microeconomic Theory by Andreu Mas-Colell, Michael Whinston, and Jerry Green. Oxford University Press, 1995.

Why Every Economist Should Learn Some Auction Theory
Advances in Economics and Econometrics: Invited Lectures to 8th World Congress of the Econometric Society, M. Dewatripont, L. Hansen and S. Turnovsky (eds.)
Cambridge University Press, 2003.

Main ideas: Here, we apply auction theory to various financial markets to illustrate the broad usefulness of framing economic problems as auctions.

Main ideas: All-pay auctions dominate other considerations, especially in personnel economics settings. These auctions are also primary tools in political economy, especially as related to rent-seeking and the study of conflict and war.

Main idea: Winning an auction may require scoring well on multiple dimensions. For instance, in procurement contexts, the winning bidder is not necessarily the lowest, but the one who offers the highest surplus. Auctions where multiple attributes combine to form a "score" are called beauty contests. We now investigate their properties.

Main idea: This topic covers information flows in organizations by examining an archetypal class of signaling games in business and public policy---corporate lobbying as cheap talk.

Special Interest Politics, Chapters 4 and 5
Gene Grossman, Elhanan Helpman
MIT Press: 2001

Main idea: This topic surveys a variety of consumer search models. The two principles play an important role in generalizing some of the findings. These models are cornerstones for modeling price competition.

"Information, Search, and Price Dispersion,"
Michael R. Baye, John Morgan, and Patrick Scholten in Handbook of Economics and Information Systems (T. Hendershott, ed.), Elsevier Press, Amsterdam, 2006

Main idea: The usual principal-agent paradigm abstracts away from the fact that other principals might compete for the same agent---and react strategically to the contracts on offer. Here we study this possibility in the context of vote buying.