Should Modern Monetary Theory inform economic policy in the crisis?

The crisis has led to renewed interest in the analytical foundations and policy implications of Modern Monetary Theory (MMT), an unorthodox approach to the management of monetary and fiscal affairs. MMT is not a safe basis for a programme for economic recovery.

Over the past few years, Modern Monetary Theory (MMT, as presented for example, by Stephanie Kelton, Randall Wray, Jim Kavanagh and Bill Mitchell) has received considerable attention in popular economic writing and has been the subject of much policy debate around the world. The current deep economic crisis has acted to re-focus many people on the analytical foundations and policy implications of MMT, which is what might be described as an unorthodox or heterodox approach to the management of monetary and fiscal affairs, as it seemingly relaxes ex ante the government’s budget constraint.

The central idea of MMT is that government expenditure can be directly funded by the issue of fiat money, or what is now more generally defined as ‘central bank money’ (monetary liabilities, both cash/currency and overnight reserves held with the central bank by banks). In short, monetary policy is put at the service of fiscal policy in achieving the desired level of aggregate spending in the economy.

This is in direct opposition to conventional monetary policies, aimed at price stability, where the central bank decides on an interest rate, or sequence of interest rates, that will achieve its inflation target through the effect of interest rates on spending in the economy. And the central bank allows the supply of central bank money to adjust to the demand. In this view, monetary policy is responsible for achieving the desired level of aggregate spending in the economy.

The reason for the recent interest in MMT is because in a deep economic crisis, monetary policy on its own may be thought to be insufficient to achieve its goals. This raises the question of the relationship between fiscal and monetary policy to which MMT provides one answer. However, it is not the only answer to the problem of how monetary and fiscal policy should work in a crisis and it brings dangers.

In this note, I outline what I understand to be the main propositions of MMT and consider them in light of a more orthodox perspective. Ultimately, I do not consider MMT to provide a safe set of axioms on which to base a monetary and fiscal programme designed to stabilise a modern economy.

Main propositions

The currency of a nation state is a liability of that nation state and sits on the central bank balance sheet. Unlike all other liabilities of the state (the consolidated national Treasury and central bank), monetary liabilities are irredeemable: they are liabilities in name only because the holder can demand their exchange for some other store of value.

MMT argues that sufficient quantities of those monetary liabilities can always be issued to meet non-monetary national debt service obligations denominated in that currency, which are just another set of liabilities but are redeemable in central bank money.

The second proposition is that any resulting over-issuance of currency, which would normally undermine the value of the currency with a sustained bout of inflation, can be controlled by the fiscal authority, which can always raise taxes to drain excess money issuance.

These two propositions are then typically combined into a statement that says the government, via the central bank, can ex ante create money to support or boost latent aggregate demand and ex post deal with any monetary excess by draining that money through prompt contractionary fiscal policy.

Conventional propositions

As it happens, the orthodox monetary and fiscal programme has similar propositions sitting in the background. Central banks do issue money in the form of currency, or electronic reserves, and the fiscal authority can adjust the level of aggregate demand in the economy to support an inflation objective, particularly when monetary policy is prevented from providing a traditional monetary stimulus through a cut in the policy rate because the policy rate is constrained near the zero lower bound (ZLB).

Strictly speaking, the binding lower bound of the policy rate is the effective lower bound (ELB), which is slightly negative (likely around -75bps) because of the carry cost of zero interest-bearing currency.

But there are a number of critical differences between the orthodox programme and MMT.

First, any shifts in the issuance of central bank money as a veil over economic activity (currency and reserves) are always consistent with an overall programme to support monetary and financial stability (see Chadha et al, 2014). Its objective is one of price stability, as an operationally independent central bank, which means matching nominal expenditure plans to overall planned real activity at a low, stable and predictable inflation rate.

Second, central bank money is not issued by the central bank to provide resources or tokens to the state. When the policy rate (Bank Rate in the UK) is above the ELB, the central bank, having decided on the appropriate level of Bank Rate to meet its inflation target, supplies currency and reserves elastically to the banking system. The banking system economises on its demand for central bank money, as those assets do not offer a good rate of return relative to lending, but will wish to meet any demand for the conversion of bank deposits into cash and for the settlement of payments (see Goodhart, 1989).

In fact, central banks typically stress that the system as a whole can always access however much central bank money it requires to meet its needs. The price of that access is Bank Rate, which, away from the ELB, is set according to a well-understood set of procedures.

Third, in a modern economy, ‘broad money’, which accounts for some 95% of the money stock, is created by banks making loans, which create both an asset and a corresponding liability for the banking sector. Broadly speaking, then the demand for central bank (also called narrow or base) money, currency and reserves, is essentially determined by overall lending to the private sector.

The quantity of lending in this case is not without limit. It is subject to three main constraints:

The demand for new loans and the affordability of the debt held by the private sector.
The limits placed on the size of banks’ balance sheets and the overall banking system by competitive pressures, risk management and regulatory demands.
Monetary policy itself setting Bank Rate (when Bank Rate is above the ELB) and influencing rates across the maturity spectrum by changing the size and composition of the central bank’s balance sheet.

Fourth, the fiscal authority starts from the premise of meeting its present value budget constraint, which means that plans are formulated ex ante so that the present value of tax revenues meets the present value of expenditures plus the value of the existing stock of public debt.

In practice, this means that in response to any increase in government expenditure, tax revenues increases are either implemented now or deferred with the help of a announced strategy or plan. In the case of deferral, the government must sell its debt to the (non-bank) private sector or sell it abroad, in order to tap the pool of private and foreign savings.

The implication is that the suggestion by proponents of MMT that expenditures need only be funded by raising taxes if they turn out to generate excessive inflation are rather like saying we will pay the mortgage only as the bailiffs arrive to evict our delinquent household.

Back to front

MMT actually inverts the timeline of orthodox theory. In an orthodox monetary economy, the private sector formulates plans, for which loans are required. The central bank sets monetary and financial conditions accordingly in line with objectives for stability (as long as Bank Rate is not constrained by the ELB). And the state provides the necessary public goods underpinned by a tax system and careful debt management.

Under MMT, the central bank provides tokens or claims for the state to deploy to meet its economic objectives without prior statements or plans for how those tokens will be met by future taxes. These claims mean that the central bank cannot control Bank Rate, as setting quantities for central bank money means that the resulting fluctuations in demand for central bank money will set interest rates if the economy is not at the ELB.

Since these claims are uncoordinated, they also seem likely to compete with private sector resource allocation decisions. Furthermore, without an articulated plan for monetary and fiscal stability, the private sector will also be unable to understand very clearly the path of the price level or of future taxes. It is thus difficult to see how the MMT framework would lead to a credible monetary and fiscal framework.

Quantitative easing, helicopter drops and MMT

In the orthodox view, money is treated as a counterpart to decisions by the private sector.

The question has often been asked as to whether money constitutes net wealth for the private sector (see, for example, Gale, 1982) and thus can be issued without constraint in the manner proposed by MMT. Indeed, central bank money, which lies outside the private sector, can be considered a net asset held by the private sector (although it is questionable as to whether it is a liability of the central bank because of its irredeemability). But the price of borrowing central bank money would be set by Bank Rate. Accordingly, when policy rates are constrained by the ELB, other ways to influence monetary and financial conditions have to be found.

In the orthodox case, the steps in a monetary-fiscal programme are that the fiscal policy-maker has made some decision to issue non-monetary domestic-currency-denominated debt. Depending on the capacity of markets to absorb this debt, the resulting bond prices may or may not be quite where the central bank wants them to be, given the constraints on policy rates. In such circumstances, government debt might be bought from the non-bank financial sector on a temporary but probably long-term basis – in what is known as quantitative easing (QE).

But note that the debt, when issued, was expected to be funded with future taxes. Debt issuance that is not funded by taxes does not have a very promising history: Sargent (1982) tells the sorry tale of the causes of four hyperinflations.

On the other hand, monetary financing of the form proposed by MMT is the direct purchase of debt by the central bank. It bypasses the transmission mechanism in the real economy and simply hands unfunded resource allocation or tokens (money) to the Treasury, which would in normal times compete with private sector allocations but may seem able superficially to be supportive when the private sector is dislocated in a crisis. The direct purchase of debt by the central bank is akin to offering the state an overdraft.

But in the monetary financing case, there may be no intention of raising taxes to meet these overdrafts. And the bonds are held permanently by the central bank with a corresponding increase in the size of its balance sheet. But they may change if the monetised fiscal stimulus (a public spending increase, say) is inflationary. And it may be inflationary if greater money holdings by the private sector or the government raised demand over capacity constraints and the central bank was not able, or expected not to be able, to respond by tightening monetary and financial conditions (Carlin and Soskice, 2015). If that were the case, then taxes would be raised (or public spending will be cut again).

If the private sector thought that these tokens were claims on real resources, then they would have some stimulatory effect on the economy (see Buiter, 2014). Indeed, if one took the view that households would always demand central bank money, were it issued in ever larger quantities, and placed a positive value on it related to the claims on output, then even an unfunded ‘helicopter drop’ or an MMT-driven policy could always be relied on to boost nominal output.

But the prospects for a stable demand for central bank money in the presence of a large or repeated deployment of a helicopter drop or the resort to MMT seem to be strictly limited if the fiscal stimulus creates excess demand for real goods and services and thus boosts inflation. And the magnitude of any stimulatory effect seems unlikely to be much larger than a more standard form of debt issuance with QE.

It is the same in the long run, even if there is no debt neutrality. Debt issuance to fund a fiscal stimulus where the debt is ultimately monetised (through QE) has the same stimulatory effect in the long run as the same size fiscal stimulus that is monetised immediately (see Harrison and Thomas, 2019). So why take the risk of unfunded expenditures and monetary expansions that may promote instability?

Related question: 'Monetary financing': is it happening and what are the dangers?

History of thought

I hear clear echoes of the ‘Real Bills Doctrine’ in the ideas of MMT. The doctrine was probably first formalised by Adam Smith and then rejected by Henry Thornton and David Ricardo, only to be set up again by Thomas Tooke and John Fullarton in the English banking controversies of the first half of the nineteenth century.

The doctrine argues that there cannot be an excess creation of bank notes. Note first that bank notes at that time were IOUs issued by a bank, any bank, used to discount bills of exchange presented by creditors who needed liquidity. The argument was that bank notes when lent against sound commercial paper must then simply correspond to real activity, or trade, financed by those bills of exchange and could not be issued to excess.

The question then is what places a constraint on the overall issuance of bank notes in this system so that there is not an excess issue in total and, if there were an excess, would it lead to a depreciation of the value of currency in proportion to the over-issuance or would holders of excess bank notes pay off their debts and so eliminate the excess supply.

The fundamental problem is that the demand for notes will vary with nominal expenditures and so if there is an over-issuance, demand will rise accordingly and the price level will become unhinged. Without a link from the money supply to some external anchor, such as a gold standard, the real bills doctrine would lead to dynamic instability in the price level.

Where can I find out more?

‘The Deficit Myth’ review: years of magical thinking: John Cochrane’s review of Stephanie Kelton’s recent book.

Warren Buffett hates it. AOC is for it. A beginner’s guide to Modern Monetary Theory: Peter Coy’s 2019 piece in Bloomberg Businessweek, providing ‘an overview of a once-fringe school of economic thought that’s suddenly of the moment’.

Modern Monetary Theory: The views of 38 top US economists polled in 2019 as part of the regular IGM panel survey run by Chicago Booth, summary here.