Collaborative Research: Cognitively Costly Decision Making by Economic Agents: Micro and Macro Implications

Duke University, Durham NC

Investigators

Abstract

This award funds research in macroeconomics. The team starts with the observation that many and possibly most economic decisions require time, attention, and careful reasoning. In other words, economic decisions require a specific kind of work called cognitive effort. However, most standard approaches used in macroeconomics assume that people can make their economic decisions instantaneously, with no effort. The research team will develop and analyze a new and tractable model of an entire economy where people, firms, and institutions find that reasoning requires costly cognitive effort. The model allows for the possibility that individuals, businesses, and policymakers may make economic decisions that are not in their long-run best interest simply because determining the best action takes too much time and effort. This is consistent with results from behavioral science. The new model may give us new tools to predict the effects of economic policies, and therefore could contribute to U.S. economic growth as well as advancing fundamental knowledge in economic science. The cognitive limitations studied in this project are not about agents imperfectly observing the relevant state variables such as interest rates, prices, and tax rates. Instead, the focus here is on agents imperfectly computing actions such as hours worked and how much to spend and save. The standard approach in macroeconomics focuses on issues created by imperfect observation and assumes a mapping derived under full rationality. The approach taken in this research is consistent with experimental evidence. Formally, the research starts with the assumption that agents observe the objective states but do not know their optimal policy function. They therefore make rational decisions about how to expend costly cognitive resources to learn about the policy function. The team will first develop a tractable static framework of individual behavior. The key emerging property is that as information accumulates about the unknown policy function, the optimal reasoning choice makes agents' effective actions state and history dependent, leading to non-linearity, endogenous persistence, and volatility clustering. The non-linearity generates both inertia to small shocks, and a salience like effect for large shocks. The second project extends the theoretical analysis to a dynamic setting where agents fully anticipate how current reasoning choices impact future utility and choices. The third project sets the mechanism in a representative agent macro model and studies its aggregate implications, where the non-linear features of the micro behavior are likely to re-surface in aggregate outcomes. There the cognitive limitations that affect individual decision rules as well as the preception of general equilibrium effects are jointly analyzed. The fourth project introduces heterogeneity and incomplete markets. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.