New economic models can help policymakers better understand the effects of their inflation-taming measures

Much about today’s inflation is not well understood. Why are some households seriously harmed while others barely feel inflation’s impact, and may even benefit? How is the battle against inflation affected by the glut of savings and government payments brought on by the pandemic? How important were pandemic-related supply shocks and the Russian invasion of Ukraine?

The evolving objectives of monetary policy further complicate our understanding of inflation. Monetary policy has long emphasized controlling inflation by stabilizing aggregate demand. But recently, central banks have broadened their objectives to include financial stability, climate and geopolitical risks, and social inclusion.

Macroeconomic models play a key role in helping to navigate this complicated landscape. Models help policymakers interpret empirical observations about the state of the economy, suggest how different policy settings will affect their objectives, and ultimately guide policy decisions. Quantitative models measure the strengths of different forces at play, helping to assess the trade-offs between competing objectives.

But traditional models ignore income and wealth inequalities and assume that what’s good for the typical consumer, as defined by the models, must be good for the broader economy.

A newly developed class of quantitative models is particularly suited to guiding central bankers across this new monetary policy territory, in which the wealth and income distributions are a central consideration. Known as HANK models, they combine heterogeneous agent models (macroeconomists’ workhorse framework for studying income and wealth distributions) with New Keynesian models (the basic framework for studying monetary policy and movements in aggregate demand).

HANK models impart new lessons about redistribution and the heterogeneous effects of monetary policy and shed new light on traditional central bank objectives of inflation control and output stabilization. Here are four broad lessons, and some preliminary thoughts, on how HANK models may illuminate our current high-inflation environment.

Lesson 1: Predicting indirect policy impacts

HANK models have taught us how monetary policy affects household consumption expenditures, both directly and indirectly. Direct channels are those that can be directly ascribed to a change in short-term policy rates, such as consumers’ decisions to postpone purchases when interest rates increase. Indirect channels arise through the impact of the policy rate on other interest rates (such as long-term bond and mortgage rates), on asset prices (such as housing and stocks), and on dividends, wages, and government taxes and transfers.

The relative size of indirect versus direct channels depends mainly on the aggregate marginal propensity to consume (MPC), which measures how much of a household’s increase in income gets spent and how much is saved. In traditional models, which try to predict the impact of monetary policy on the typical consumer, the MPC is tiny, and consequently the indirect channels are insignificant. HANK models, instead, are built to be consistent with empirical evidence on consumption and saving behavior. Their aggregate MPC is roughly 10 times larger, and thus the various indirect effects dominate the transmission mechanism.

What does this mean for monetary policy? Through the lens of older models, all a central banker needs to know to predict the aggregate consumption response is an estimate of one parameter, consumer willingness to postpone purchases when interest rates rise (the “intertemporal elasticity of substitution”). But with HANK models, central banks need much more exacting information about the household side of the economy. They need a full picture of the distribution of MPCs, income sources, and the components of household balance sheets. In addition, the importance of indirect channels means that the transmission of monetary policy is mediated by all those mechanisms that contribute to price formation in goods, inputs, credit, housing, and financial markets. Therefore, central banks need a deep comprehension of market structures and frictions, as well as of institutions that play major roles in these settings, such as local governments, unions, and regulatory bodies.

Lesson 2: Some ships are lifted higher, others are sunk

In the traditional view of monetary policy, “a rising tide raises all ships.” HANK models show this is a fiction.

Many channels of monetary policy have divergent, and sometimes opposing, effects on different households. For example, the direct effects of interest rate changes depend on households’ balance sheets: rate cuts benefit debtors, whose interest payments decrease (such as households with adjustable-rate mortgages) and hurt savers, whose interest income falls. Monetary policy also has heterogeneous effects through its impact on inflation. First, inflation benefits households with lots of nominal debt that is revalued downward. Second, prices rise more for some goods than for others, and different households consume these goods in unequal proportions. Finally, the indirect effects of monetary policy on household disposable income are uneven because some households are more exposed to fluctuations in aggregate economic activity than others.

In HANK models, these redistributive channels are not only crucial for understanding who wins and who loses in monetary policy but are also at the core of how monetary policy operates, in the sense that redistribution determines its quantitative effect on macroeconomic aggregates. To the extent that the channels outlined above redistribute from households with low MPCs to those with high MPCs (from savers to spenders), the macroeconomic impact of monetary policy is amplified. These redistributive effects will also differ across countries. For example, they would likely be stronger in countries with a high poverty rate or high inequality, thereby also resulting in different monetary transmission between advanced economies and low- and middle-income countries. HANK models force us to let go of the fiction that we can cleanly separate stabilization from redistribution.

Lesson 3: Fiscal footprints matter

Another widespread misconception is the view that monetary policy can be divorced from fiscal policy.

By introducing income and wealth inequality, HANK models reestablish a strong link between the two, showing how monetary policy leaves consequential “fiscal footprints.” When the central bank raises interest rates, the treasury’s borrowing costs increase, and the increase must be funded by raising taxes or lowering expenditures, now or in the future, or through future inflation. In HANK models, the details of how and when the government makes up this fiscal shortfall, and which households bear the burden, have a tremendous influence on the overall effects of interest rate hikes.

The fiscal footprint of monetary policy thus generates additional redistribution, which, in turn, amplifies or dampens the shock depending on whether it shifts resources from savers to spenders, or vice versa. This force keeps central banks and treasuries inseparably intertwined. The more debt the government owes and the shorter-term it is, the bigger the fiscal footprint.

More generally, HANK models are also a natural environment to study the effects of fiscal policy on aggregate productive efficiency, the degree of social insurance, and the extent of redistribution between households.

Lesson 4: The right tool for redistribution

Where does this leave monetary policy in practice?

Studies of optimal monetary and fiscal policy in HANK models agree that the benefits of aggregate stabilization are dwarfed by the gains from directly alleviating hardship. Optimal policies in HANK models almost always favor redistributing toward hand-to-mouth households in downturns.

One may be tempted to read this as endorsement of using monetary policy to share prosperity and mitigate adversities. But monetary policy is a blunt tool for redistribution or insurance. HANK models tell us that fiscal policy is likely better suited for this task because it can be targeted more precisely to those in need of support.

The current bout of inflation

The current inflation episode is a good example to explore where HANK models can be useful for macroeconomic analysis and policy advice.

HANK models show that the impact of a macroeconomic shock on aggregate spending is larger when individual MPCs and individual exposures to the shock are more strongly correlated. In the current economic environment, this means that understanding the redistributive implications of inflation across households is crucial to gauge its aggregate implications. Households consume different bundles of goods and services, making some more sensitive to inflation than others. For example, poor families who spend a larger share of their income on basic goods like energy are especially harmed in this episode. Borrowers gain as the real value of their debt falls, while households with large amounts of cash or liquid savings lose. Workers whose compensation is relatively flexible (for example, because of bonuses and commissions) can limit their loss of purchasing power, whereas workers whose nominal wages are negotiated infrequently, or those paid the minimum wage, will see their real earnings shrink.

The level of household savings, which influences how a change in interest rates affects consumption, is critical. So is the distribution of savings across the population and the correlation with household willingness to spend. For example, excess savings that arose from consumption restrictions due to the pandemic (think of less spending on travel and restaurant meals) are largely held by the well-off and are therefore being spent at a very low rate. Excess savings accumulated from the large government transfer programs in 2020 and 2021 are largely held by low-income households and are being spent much faster. A rapid spending rate sustains aggregate demand and gets in the way of a central bank’s efforts to tame inflation.

Finally, a full evaluation of the welfare effects of the current bout of inflation cannot ignore its causes. The jury is still out on the relative importance of supply shocks (due to the COVID-19 pandemic and war in Ukraine), the large fiscal stimulus in 2020 and 2021, and the loose monetary policy in the decade since the most recent recession. Each of these factors had redistributive components and heterogeneous effects that cannot be understood within the shackles of traditional models. Putting HANK models to work will help us understand the full effects of this episode of monetary history.

Opinions expressed in articles and other materials are those of the authors; they do not necessarily reflect IMF policy.