Disaggregating macroeconomics

The Institute of Economics will hold a meeting of its Seminar Series on Tuesday, February, 2020: Andrea Teglio, from the Ca' Foscari University of Venice, will present the paper "Disaggregating Macroeconomics".

Abstract:

The aim of this paper is to study when, why and under which conditions a representative agent model is able to correctly approximate the output of a more realistic model based on the true assumption of many interacting agents populating the economy. The simple and widespread Keynesian cross diagram textbook model is used as the baseline representative agent model. Then, the model is decomposed in an agent-based network model, populated by households and rms which collapses on the original Keynesian cross diagram when the number of agents is one per type. This disaggregation" exercise allows us to understand when the representative agent model is no more a good representation of what's going on in the economy. Results show that the representative Keynesian cross diagram model stops being a good approximation when (i) the network structure of the economy is not symmetric enough, e.g. rms have dierent sizes, (ii) the rationality of the agents is not high enough, e.g. households are not able to decide where to purchase goods, and in particular for a combination of (i) and (ii). Therefore, the network structure of the economy and the degree of rationality of the agents (along with their information set) play a decisive role for understanding if and when it could be convenient to use a representative agent approximation.