Economic Fluctuations, Heterogeneity and Macroeconomic Policies

Project overview

The main goal of this project is to advance the knowledge that exists on the implications of heterogeneity for macroeconomic policies. In particular, the research has focused on a class of macroeconomic models that have emerged in recent years, which are often referred to as Heterogeneous Agent New Keynesian (HANK) models, and which are gradually becoming the new reference framework for the study of economic fluctuations. Models in that category feature a realistic heterogeneity in households’ income and wealth, resulting from the presence of idiosyncratic shocks to individual income, and the impossibility to insure against those shocks in the financial markets. This project has focused on developing a  tractable theoretical framework to understand and quantify the role played by household heterogeneity for the transmission of aggregate shocks (e.g. monetary policy shocks).

Summary, output, and dissemination

Research summary

The main results of the research findings in this project can be summarized in four different parts.

Result 1: Development of a simple “organizing” framework 

The most relevant contribution to research and innovation from this project has been the development of a simple “organizing” framework to understand and quantify the implications of household heterogeneity for aggregate economic fluctuations. Using this framework, it has been possible to identify four components (or channels) through which household heterogeneity might affect the transmission of aggregate shocks to aggregate variables. The framework is based on the idea of dividing households into two categories: those who are financially constrained and those who are unconstrained. Within that framework it is shown that one can derive an aggregate demand relationship, which shows how heterogeneity, affect aggregate consumption. Also, it is shown how the impact of heterogeneity can be decomposed into four components pertaining respectively to (i) the average consumption gap between constrained and unconstrained households, (ii) the share of constrained households, (iii) the risk that a currently unconstrained household become constrained in the future (labeled as “switching” component) and (iv) the dispersion within unconstrained households, as a consequence of heterogeneity in income, wealth, etc. The joint response of those components to aggregate shocks ultimately determines the differential behavior of a given heterogeneous-agent economy, relative to a stylized representative agent benchmark.

Result 2: Quantitative assessment of the role of heterogeneity

The results from the development of the “organizing” framework, have been used to assess the importance of heterogeneity from a quantitative viewpoint. The chart in Figure 1 titled “Impulse Responses” shows the response of the four components of heterogeneity to a 1% reduction in interest rates. The total effect of heterogeneity (represented by the red line) is an increase in aggregate consumption of 0.3% above its trend. However, such an increase results from the combination of different forces (blue, yellow and purple lines) some of which go in opposite directions. 

This project shows that while the consumption “gap” between unconstrained and constrained households (blue line) is pro-cyclical (i.e. inequality decreases during in good times), the risk of becoming constrained (purple line) is countercyclical. Also, the “dispersion” within the unconstrained households (green line) is largely acyclical, suggesting that such a dimension of heterogeneity plays a minor role in the propagation of aggregate shocks to economic aggregates.

Figure 1

Result 3: A tractable alternative to HANK 

The researchers of this project were able to develop a simple heterogeneous agent framework that admits an analytical (paper and pencil) solution, which is suitable to be included in medium- or large-scale models and used by central banks and other policy institutions. This framework features two types of households, constrained and unconstrained (or employed and unemployed workers). In each period, households are affected by idiosyncratic shocks to their income and may become constrained (or unemployed) at a future date. The research showed how such a simple framework captures three of the four components of heterogeneity: “gap”, “share” and “switching”. The fourth component, the “dispersion” within unconstrained households, would only arise in a more complex model that explicitly takes into account heterogeneity in households’ wealth and financial portfolios, as in the workhorse HANK model, and how such heterogeneity evolves over time. 

A natural question is to what extent omitting these features limits our understanding of the economic fluctuations, and this is exactly what can be analyzed within the framework of this project. 

Figure 2 shows how a simple two-agent alternative (represented by the blue line) tracks almost perfectly the behavior of aggregate output implied by a fully-fledged HANK model (represented by the red line) in response to a monetary shock. 

Figure 2

This suggests that a simple alternative may provide a good approximation, both from a quantitative and qualitative viewpoint, of the effects of heterogeneity on economic aggregates. The researchers are currently investigating the robustness of these findings in alternative environments.

Prizes and Recognition

• The Wim Duisenberg Fellowship was awarded to Davide Debortoli by the European Central Bank, 2019
• Related project “Heterogeneity, Monetary Policy and Economic Fluctuations (HEMPEF)” has been awarded a Senior ERC grant to Jordi Galí

Conferences and workshops

Conferences

• The Heterogeneous Effects of Macro Shocks and Policies, organized by Norges Bank, November 2022
• Paris School of Economics, Fall 2022
• Impulse and Propagation Mechanisms, organized by the National Bureau of Economic Research (NBER), July 2022
• European Central Bank, Spring 2022
• The University of Oxford, Spring 2022
• Univ. of Lausanne, Winter 2022
• Bocconi University, Fall 2021
• University of Maryland, Fall 2021
• London Business School, London, Fall 2021
• Ecole Polytechnique – CREST, Paris, Spring 2021
• Bank of Canada, Ottawa, Spring 2021
• Boston College, Boston, Spring 2021
• University of Maryland, Spring 2022

Workshops

• Copenhagen Macro Workshop, Fall 2020
• Bundesbank DSGE Macro Workshop, Frankfurt, October 2019 
• ECB/CEPR Workshop “Household Heterogeneity in Macroeconomics,” Frankfurt, December 2018