Work in Progress
The Cost of Heterogeneity: Can Density Peaks Clustering Improve CES Aggregation? (Job Market Paper)
As datasets have become larger and more widely available, researchers across many fields are rapidly developing new tools for data analysis. This paper applies recent advances in data science to a long-standing macroeconomics question by using density peaks clustering to improve estimation of aggregate price indexes. In simulations, I show that the algorithm can identify groups of consumers whose purchases are generated from a logit demand system with heterogeneous tastes. Clustering consumers in this way can make the assumptions needed to consistently estimate elasticities more plausible. Applying the method to a panel of consumer retail purchases, I show that standard methods on unclustered data may overstate the degree of product substitution.
Keynesian Dynamics with Customer Markets in a Visual, Interactive Agent Based Model
This paper adds visual and interactive elements to an agent based Keynesian economy. While the primary drivers of the economy are Keynesian inventory adjustments, the model is also centered around the premise of customer markets as firms compete through prices to retain their customer base. Within this framework, this paper describes the functionality of an accompanying standalone program that allows users to adjust parameters and observe the dynamics of the economy in real time. A benchmark calibration of the model demonstrates that the economy can generate endogenous cycles in key economic variables.
Rationality Without Rational Expectations: Alternative Methods for Modeling Expectations in Macroeconomics
The rational expectations hypothesis has been criticized both for the strong assumptions it requires of agents in economic models as well as the results implied by those assumptions. Answering these concerns, a wide class of models has been developed that allows agents to form beliefs based on misspecified models of the economy. In this paper, I explore many of these alternatives to rational expectations. In particular, I focus on two widely used methodologies: the adaptive learning of Evans and Honkapohja (2001b) and the evolutionary selection of Brock and Hommes (1997). After introducing these frameworks in a simple cobweb model, I provide a survey of many extensions and applications to more complete models that have been developed in the time since these models were introduced. I also offer a discussion of the future of research in this area.
The Market Socialism Misunderstanding: The Impact of the 20th Century Socialism Debate on Modern Economics
I argue that many of the lessons from the market socialism debate were not absorbed by modern economics. In particular, I show that Mises’s and Hayek’s argument against socialist planning stems from the importance of entrepreneurs and the dynamic market process, features that remain absent from current models.