You are in: LeeX » Events » BLEESS-M-2009 » Course Lectures

Barcelona LeeX Experimental Economics Summer School in Macroeconomics in Universitat Pompeu Fabra.
June 15-19, 2009: Lectures

John Duffy | Frank Heinemann | Rosemarie Nagel

John Duffy

• Overview of Macroeconomic Experiments

This lecture will expose participants to the breadth of macroeconomic topics and questions that have been explored using laboratory methods. The aim of this lecture will be to stimulate thinking about ideas for new projects that build on what has already been done. In addition, participants will be encouraged to extend laboratory methods to macroeconomic models or questions that have not been previously addressed. Methodological issues that are particularly relevant to macroeconomic experiments, e.g., implementation of discounting and infinite horizons, will also be addressed.

• Readings:

• Asset Pricing: Bubbles, Crashes and Expectations

Currently, economies around the world are experiencing an economic downturn brought about by the collapse of housing and equity prices and the deleveraging of the financial institutions that underwrote those assets.  In this lecture we examine laboratory studies addressing asset pricing and the phenomenon of asset price bubbles and crashes.  An understanding of the causes of asset price bubbles and cashes is of obvious importance to both policymakers and asset market participants. While there exists experimental designs that reliably yield asset price bubbles and crashes among inexperienced subjects, there remains much more work to be done on this topic, for instance, there is a need for an experimental design in which asset price bubbles and crashes are recurrent phenomena.

Smith, Vernon, Gerry L. Suchanek and Arlington W. Williams, 1988. “Bubbles, Crashes, and Endogenous Expectations in Experimental Spot Asset Markets,” Econometrica, 56, 1119-1151. Lei, Vivian, Charles N. Noussair and Charles R. Plott 2001. “Nonspeculative Bubbles in Experimental Asset Markets: Lack of Common Knowledge of Rationality vs. Actual Irrationality,” Econometrica, 69, 831-859. Dufwenberg, Martin, Tobias Lindqvist and Evan Moore, 2005. “Bubbles and Experience: An Experiment,” American Economic Review, 95, 1731–1737. Hommes, Cars.H., Joep Sonnemans, Jan Tuinstra and Henk van de Velden, 2005. “Coordination of Expectations in Asset Pricing Experiments,” Review of Financial Studies 18, 955-980. Ernan Haruvy, Yaron Lahav and Charles N. Noussair, 2007. “Traders' Expectations in Asset Markets: Experimental Evidence,” American Economic Review 97, 1901-1920. Crockett, Sean and John Duffy, 2009. “A General Equilibrium Approach to Asset Pricing Experiments.” working paper.

• Monetary Theory

Among the central questions in monetary theory are why intrinsically worthless fiat money serves as a store of value and why it is used as a medium of exchange when other assets dominate it in rate of return. Various theories have been developed to address these fundamental questions. For instance, overlapping generations models of money may explain why fiat money has value, and search-theoretic approaches can rationalize why money is used when dominated in rate of return by other competing assets. However, the frictions in these models -overlapping generations and search frictions- make them difficult to take to field data. On the other hand, a number of laboratory studies of such models have been conducted. These lectures will outline the main findings from those studies and point out promising new extensions.

• Literature:

Introduction:

Theory:

Experiments:

• Sunspots

William Stanley Jevons, the Victorian era economist/polymath who helped launch the marginalist revolution, believed that the solar cycle drove the business cycle and he collected much data in support of this theory. As it turns out, business cycles are a lot more irregular than the 11-year sunspot cycle and there is considerable variation in the timing and duration of business cycles across countries. So today, we honor Jevon's folly by referring to nonfundamental or extraneous factors that may affect economic activity as "sunspots" (also "animal spirits, "self-fulfilling prophecies") Examples include changes in the length of women's hemlines, or, in the U.S., whether a team from the National Football Conference wins the Super Bowl or the economic predictions of the Wall Street Journal.

Formal, elegant models in which sunspot variables matter in a rational expectations equilibrium are found in the seminal work of Cass and Shell (1983) and Azariadis (1981). A difficulty with testing sunspot theories concerns identification of the non-fundamental variable agents may be coordinating and conditioning upon. Laboratory methods can be helpful in this regard, and in this lecture we review a couple of experimental studies that have sought to demonstrate the existence of an equilibrium in which sunspots matter.

Theory:

Experiments:

Frank Heinemann

• Speculative Attacks and the Theory of Global Games - Experimental Tests of Global Game Predictions

Speculative attacks can be viewed upon as coordination games: if a sufficient number of traders (and a sufficient amount of capital) is involved in an attack, the pressure on foreign exchange markets forces the central bank to devaluate its currency. Then, all attacking traders gain from the devaluation. But, if the number of attackers is too small, the central bank can defend the peg, and attacking traders lose on transaction costs. Speculative-attack games have multiple equilibria if payoff functions are common knowledge. The theory of global embeds a coordination game in an environment with private information about parameters of the payoff function. If private information is sufficiently precise, the global game has a unique equilibrium. Hence, the theory of global games can be used for a unique prediction of the outcome of a speculative-attack game. This theory provides a number of hypotheses that can be tested in laboratory experiments. The lecture on Speculative Attacks and the Theory of Global Games presents some of the theoretical background and derives testable hypotheses. The lecture on Experimental Tests of Global Game Predictions explains experiments that have been used for these tests and shows how they have been analyzed.

• Literature:

Introduction:

Theory:

Experiments:

• Monetary Policy and Expectations

In the 1960s and early 1970s, policymakers thought that they could systematically raise employment by inflationary monetary policy. Kydland and Prescott (1977) and Barro and Gordon (1983) have shown that an asymmetric objective function of the central bank gives rise to an inflation bias stemming from a time inconsistency: ex ante, the central bank would like to commit to a low average rate of inflation. Ex post, however, after expectations have been formed, the asymmetric objective function provides an incentive to deviate from such a commitment and raise inflation in order to stimulate the economy and raise the output level.
In a rational expectations equilibrium this incentive is expected and private agents expect a rate of inflation above the social optimum, at which the central bank has no incentive for a further rise of inflation. Laboratory experiments can test the hypotheses of time inconsistent policy as well as rationality of expectations in such an environment. The lecture on Monetary Policy and Expectations will present the fundamental problem of time inconsistency and an experiment designed to test this theory.

• Literature:

Theory:

Experiment:

• Stabilizing Inflation and Employment

Dynamic stochastic general equilibrium (DSGE) models are the new paradigm of macroeconomics. Within these models, monetary policy is reduced to a dynamic control problem. Under some restrictions, the optimal solution to such a control problem can be described by a simple linear function with which the instrument (interest rate) responds to current (and lagged) inflation and output. One of the major problems of modern central banks is model uncertainty. There is uncertainty about the true model and its parameters. With model uncertainty, the instrument should gradually respond to observed deviations of inflation and output from their target values. In reality, monetary policy is conducted by committees instead of single agents. This raises the questions whether and why groups are better in dealing with dynamic control problems than individuals and how they come to an agreement in cases of diverging opinions. Some recent experiments put subjects in the role of central bankers and let them solve dynamic control problems of an economic model whose parameters are not known to subjects. Subjects have to solve these problems either individually or ion small groups. It turns out that most subjects are able to control inflation and their strategies can be described by Taylor-type rules. Groups do significantly better than individuals. The modes by which groups are organized (size, voting rules, leadership) have no significant effect.

• Literature:

Theory:

Experiment:

Rosemarie Nagel

• Methodology

This lecture introduces the methods of experimental economics. We will discuss what is an economics experiment, why we do experiments, the different areas in experimental economics and behavioral economics, the link between experimental economics, theory and empirical work and important design issues.

• Literature:

• Rational and boundedly rational expectation: a micro and macro view

In this lecture we will discuss how micro and macro economics incooporate bounded rational expectations in their models. Most macroeconomic models presume that agents (usually a representative agent) posses rational expectations, that is, expectations consistent with knowledge of the model. Laboratory tests of the rational expectations hypothesis suggest that subjects do not initially form expectations consistent with the rational expectations hypothesis; in many cases, expectations are more consistent with an adaptive learning process. However, there is some evidence that individuals can learn to form rational expectations with sufficient experience.

Most micro economic models assume that people make acurate predictions about others behavior. Typically one assumes common knowlege of rationality among the agents which means that everybody is rational and thinks that all are rational etc. Based on this assumption equilibria are calculated. There are at least two streams to explain deviations from equilibrium. One part is to develop learning models which deviate in different degrees from rationality. The second stream is to assume that an agent might think that he is rational but does not think that others are rational but instead models them as random players. Or he might think that others are rational but these others do not believe that their coplayers are rational. etc. This way one can formulate different degrees of levels of reasoning about the rationality of others. We will discuss this second stream of literature.

• Literature: