RDP 2025-04: HANK and the Transmission of Shocks to Demand and Supply 1. Introduction

In this paper, we study the importance of heterogeneity and individual microeconomic behaviour for aggregate outcomes in the Australian macroeconomy. Specifically, we make a first attempt at considering how differences in productivity and income across individuals in Australia can lead to differences in their decisions about consumption, savings, and working hours. Further, we explore how shocks to the economy impact different individuals in different ways, and how their reactions to these shocks may matter for the overall response of the economy.

Over recent decades, there's been an increasing realisation that distributional aspects can be important for the transmission of shocks to the economy, as well as the distributional impacts of policies and shocks. The heterogeneous agent New Keynesian (HANK) approach provides a powerful tool to better understand the transmission of policy at the household level by microfounding the optimal decisions of individuals with diverse outcomes. It allows for the understanding of how heterogeneity can change the the transmission of aggregate economic shocks, as well as how such shocks affect inequality (Acharya and Dogra 2018).

There are a number of channels through which a nontrivial distribution of income and wealth may affect shock transmission as compared to a world with a representative agent. For example, in a representative agent setting, all individuals are net savers, and will experience a fall in interest income and a rise in wage income in response to a fall in interest rates, with the net effect being an increase in consumption due to intertemporal substitution. In a heterogeneous agent setting, some households will be net savers while others will be net borrowers, and they will experience different effects on their income sources depending on where they land in the distribution. Depending on how a given household's current and expected future income is impacted, they may increase consumption in response to a rate cut while others may reduce it, with a similar response for labour hours and asset holdings. These variable responses could change the overall impact of the shock on the economy.

In turn, there are also channels through which the shock may affect the distribution of wealth. Leong (2021) highlights several possible channels through which monetary policy easing can affect wealth inequality. For example, there is the savings redistribution channel, whereby lower interest rates decrease fixed income returns and payments on debt, making borrowers better off and savers worse off. Similarly, there is the portfolio composition channel, wherein lower interest rates increase asset prices, possibly making the holders of these assets wealthier. While these channels suggest that policy shocks may have an impact on the wealth distribution, they likely act in the opposite direction for rate increases, suggesting that monetary policy may be neutral with respect to the distribution over the business cycle.

The primary aims of this paper are to examine the impact and transmission of aggregate economic shocks in a basic HANK model calibrated to the Australian economy. While much of the literature examines such models featuring multiple asset classes and unforeseeable economic shocks, in this work we consider a single asset, government bonds, which is closer in spirit to traditional representative agent New Keynesian (RANK) models. We do this as a first step towards building more sophisticated models for the Australian economy.

While this study is an important first step towards modelling the implications of heterogeneity for the transmission of shocks and conduct of monetary policy in Australia, these initial results are subject to a number of caveats. Shock transmission may be affected by the details of the composition of household assets and liabilities. This may be particularly important given the prevalence of less liquid assets such as housing and superannuation in Australia. Likewise, we do not consider secondary markets for bond assets in our economy, wherein nominal bond values could fluctuate for reasons other than interest rate changes. Moreover, we do not consider other aspects of heterogeneity which may be important for shock transmission, such as the life cycle or unemployment. Finally, in order to focus on realistic income and wealth dynamics without additional complications, the current study does not consider open economy dynamics. Further modelling in Australia will be needed to fill these gaps.

Having constructed the model, we proceed to describe various supply and demand shocks that could be considered in the model and compare the response of the heterogeneous agent economy to that of a representative agent counterpart. On the demand side, we consider a surprise decrease in the nominal interest rate (a monetary policy shock). We find that heterogeneity dampens the immediate real impacts of this shock, but slightly amplifies the inflation impacts. The dampened real response is in disagreement with much of the international literature on heterogeneous agent models, and likely reflects the lack of liquidity-constrained, high marginal propensity-to-consume households.

We then examine how household behaviour varies across the distribution of wealth in response to the monetary policy shock – by examining how the optimal consumption, work, and savings decisions adjust on impact. We find divergences in this behaviour across the distribution; for example, increased consumption by the middle quintiles of the wealth distribution is offset by those agents at the extremes. More specifically, the decline in interest rates results in a loss of income for very wealthy households who hold large stocks of liquid assets, causing these agents to cut consumption. On the other hand, the poorest households respond to a resulting increase in real wages by working longer hours, using the additional income to move away from the borrowing constraint. Correspondingly, we observe that the monetary policy easing generates a decline in wealth inequality in the model as measured by the Gini coefficient.

Next, we turn to shocks on the supply side of the economy. In particular, we consider two types of such shocks. First, we examine a shock to households' labour supply which increases their disutility from working. Second, we examine a shock to the substitutability between firm inputs (a mark-up shock). In the case of the former, we once again find considerable variation in behavioural adjustments across the wealth distribution, while in response to the latter we find much more homogeneity. This suggests that heterogeneity may be more important for shocks which directly impact the household optimisation problem. In both cases, we once again find dampened effects on the real economy.

The rest of the paper is organised as follows. Section 2 describes the literature. Section 3 outlines the model. Section 4 describes the data and calibration procedure, Section 5 outlines the shocks considered in the model, Section 6 presents our quantitative results, and Section 7 concludes.