Author: Ben Ashby

In theoretical physics, the concept of parallel universes suggests that multiple universes coexist with slightly different laws of physics and histories. I feel much the same way about Modern Monetary Theory (MMT). In this infinite array of universes, there is likely to be one where humans, financial markets, economics, and, indeed, time operate in radically different ways, and MMT could work as intended. Unfortunately, this isn’t the universe we inhabit.

I am writing this as I was at a financial conference recently, where the moderator – much to my surprise - was a strong advocate of MMT. It also seems to be oozing its way back into public discourse as it appears to have become a central part of the UK’s Green Party’s economic proposition, and based on current polling, they might form part of the next UK government.

As tempting as it is to immediately dismiss MMT it does contain genuine insights about how sovereign money actually works. It also prescribes the precise policy regime that, across five thousand years of monetary history, has most reliably destroyed the coordinating mechanisms that capitalism depends on. The two facts are not unrelated.

What MMT claims (and gets right)

Modern Monetary Theory is frequently caricatured as the argument that governments can print money without limit. That is not what its serious proponents - Randall Wray, Stephanie Kelton, Warren Mosler, and before them, Abba Lerner's functional finance - actually say. The core propositions are more carefully stated and, in important respects, descriptively accurate.

The first proposition is operational: a government that issues its own non-convertible, floating-rate currency cannot be forced into involuntary default on obligations denominated in that currency. It can always create money to meet domestic commitments. This is not a theory - it is a statement of institutional fact that applies to the United States, the United Kingdom or Japan, for example. It emphatically does not apply to eurozone members (whose debts are in a currency they do not control), dollarised economies, or any issuer with significant foreign-currency debt. The distinction matters enormously, and MMT is correct to insist on it.

The second proposition draws on Wynne Godley's sectoral balances framework: the financial surplus of one sector must be the deficit of another. If the private and external sectors both wish to net save, which is the normal condition for a current account deficit country, then the government sector must run a deficit as a matter of accounting identity. This is not ideology; it is a closed-system constraint. Attempts to achieve simultaneous private deleveraging, external surpluses, and government austerity violate basic accounting principles and produce exactly the recessions that Godley predicted in the 1990s; the ‘recoveries’ of some periphery countries since the Eurocrisis are a good example of that.

Chart: When you have done 6 impossible things before breakfast; Greece’s post crisis GDP

 
Source: Eurostat & IMF

The third proposition is that taxation does not finance government spending in a monetary sense. Rather, spending creates the money that taxation subsequently destroys, with the purpose of taxation being to drain aggregate demand (and to give the currency value by creating a compulsory liability payable only in state money). Government bonds, in this view, are not debt instruments that fund the state; they are a mechanism for draining excess reserves and setting a floor on short-term interest rates. This is a sophisticated restatement of Lerner's functional finance, and it is not, in purely monetary-plumbing terms, wrong.

Where it actually works 

The MMT framework performs well in three specific settings. First, in liquidity trap environments - Japan being the clearest example - it correctly identifies that deficit spending need not crowd out private investment or produce inflation when the private sector is deeply risk-averse and the central bank accommodates. Japan has run debt-to-GDP ratios well in excess of 200% for decades without the hyperinflation or bond market collapse that conventional models perpetually forecast. The MMT observation that a currency-issuing sovereign with a domestic savings surplus faces a different constraint set from that of a household or a euro member is vindicated there.

Second, MMT correctly diagnoses the political economy of austerity. The idea that government must "balance its books" like a household is not merely economically confused - it actively misleads policymakers into procyclical fiscal tightening. The UK's experience between 2010 and 2016, during which sustained demand suppression deepened the output gap and paradoxically slowed deficit reduction, is consistent with the MMT critique. So is the IMF's own subsequent research acknowledging that fiscal multipliers in low-rate environments exceed one.

Third, the Job Guarantee proposal - often flagship policy of MMTers - has theoretical coherence. The idea is to use the government as the “employer of last resort”, creating a buffer stock of employed workers rather than the current buffer stock of unemployed ones. Since the wage paid sets the price floor (rather than the quantity of money supply), the proposal claims to be non-inflationary by construction. This is at least theoretically serious, even if implementation at the required scale and administrative sophistication has never been demonstrated, and in my opinion is extremely unlikely.

And why it doesn’t withstand any contact with this reality

In practice, however, three problems present themselves that the theory waves away with remarkable insouciance. 

MMT substitutes inflation for solvency as its operative limit. Spend freely until consumer prices signal resource exhaustion, then raise taxes to drain demand. Simple, elegant, and dependent on institutional assumptions so heroic they belong in a different universe inhabited by a different type of human.

I have previously written about the structural problems in measuring inflation, but MMT's difficulties go considerably deeper than that. [i]

What it does not adequately theorise is the political economy of the constraint itself. In practice, democratic governments facing an inflation constraint do not tighten with the clinical precision that theory requires. They delay, they deny, and they blame supply chains. The institutional assumption buried inside MMT - that the fiscal authority will respond to inflation signals as reliably as a well-functioning central bank would respond to solvency signals - is heroic to the point of being unrealistic. The post-2020 experience in the United States, where the American Rescue Plan contributed to an inflationary surge that proved far more persistent than the Treasury or the Federal Reserve initially acknowledged, is a cautionary instance.

The inflation problem is compounded by what might be called the credibility loop. The value of a fiat currency rests on the expectation that the issuer will not debase it without limit. Central bank independence was constructed precisely to provide a credible institutional separation between monetary financing and fiscal spending. MMT, taken to its logical conclusion, collapses this separation.

The external sector (and the financial sector) is the second major weakness. MMT is largely a closed-economy model. It handles the current account only through the sectoral balance's identity, which tells you the accounting relationship but not the behavioural dynamics. A persistent fiscal deficit that is monetised will tend, ceteris paribus, to depreciate the exchange rate. 

For the United States, which issues the world's primary reserve currency and faces near-unlimited demand for its paper, this constraint is genuinely looser than for any other economy. The dollar's exorbitant privilege - the ability to export inflation and run structural deficits that the rest of the world finances - is a real phenomenon, and some MMT insights are more applicable to the US than anywhere else. For the United Kingdom, whose currency reserve status is modest and whose current account deficit is structurally large, the margin for sustained monetary financing is correspondingly narrower. The gilt market's violent reaction in September 2022 to an unfunded fiscal expansion - even one that did not reach the scale MMT might contemplate – is instructive.

The act of visibly financing deficits through money creation can trigger a confidence crisis - a self-fulfilling loss of currency credibility - that arises well before any real resource constraint. The exchange rate is the transmission mechanism. A currency that loses credibility faces import price inflation, a vicious depreciation-inflation spiral, and potentially a sudden stop in capital flows. Argentina has visited this sequence repeatedly. Turkey visited it between 2021 and 2023, when the Erdoğan administration effectively ran a version of MMT-adjacent logic by suppressing interest rates despite rising inflation. The outcome was a lira that lost over 40% of its value in a matter of months.

Chart: Turkey’s experiences with money should be a warning

 
Source: Bloomberg

The interest rate channel represents a third problem. MMT's operational story holds that the central bank sets the short rate and fiscal deficits add net financial assets to the private sector, with bond issuance simply swapping one form of government liability for another. This is broadly correct in the short run and at low levels of deficit. But at scale and over time, the term premium is not simply a policy variable. Investors hold long-dated government bonds in competition with every other claim on future purchasing power.

If deficit monetisation becomes expected and persistent, term premia will rise to compensate for inflation risk, raising the cost of all long-duration borrowing in the economy - mortgages, corporate bonds, infrastructure financing - and potentially triggering the crowding-out that MMT claims is a myth.

Finally, and perhaps most fundamentally, MMT underweights the supply side. The binding real constraint is productive capacity - the combination of capital, labour, technology, and institutional quality that determines how much an economy can actually produce without inflation. Demand management, whether through fiscal or monetary tools, cannot create productive capacity; it can only help utilise existing capacity more or less fully. The correct insight that idle capacity can be put to work without inflation tips too easily into the dangerous assumption that there is always idle capacity available - or that the capacity lost through structural change, deindustrialisation, or demographic shift can be conjured back via spending.

It cannot. 

I guess we are not in Japan anymore, Toto

Japan is the MMT exhibit that never quite goes away, and it deserves a serious answer rather than a dismissive wave.

For three decades, Japan ran debt-to-GDP ratios that would have caused many a conventional economist's model to implode, held rates at zero, and got deflation rather than the hyperinflation or bond market vigilantes that every sensible person kept predicting would arrive any quarter now. The MMT crowd point to this, not unreasonably, as evidence that the standard solvency narrative is wrong.

They are not wrong about Japan. That is precisely the problem.

Japan worked - in the specific, narrow, non-transferable sense - because it is a surplus creditor nation whose domestic savings base absorbs its own sovereign paper without needing to persuade anyone external to show up. Over 90% of Japanese Government Bonds are held domestically by investors who are culturally and structurally disinclined to go elsewhere. This is not a monetary policy achievement. It is a demographic and institutional inheritance that no other major economy shares or can plausibly replicate.

And even Japan, it turns out, is not Japan anymore. Inflation has returned above 3%, wages are accelerating, and the Bank of Japan has begun raising rates from zero. The escape valve is closing. The MMT crowd have been unusually quiet about this.

More importantly, Japan is simultaneously the most sustained case study of what permanent rate suppression actually produces over time: three decades of stagnation, zombie bank proliferation, a corporate sector that hoarded cash rather than invested it — because the price of capital gave no honest signal about productive opportunity — and a demographic retreat from productive risk-taking that makes the fiscal position look, with each passing year, rather less comfortable than the headline numbers suggest. MMT claims Japan as vindication. The prosecution would like to submit it as Exhibit A.

The broader point is simple. Japan is not the UK, the United States, or any economy that relies on external financing and lacks a captive domestic savings base. The liquidity trap conditions that made deficit spending genuinely safe in 2009 are not permanent features of the economic landscape. They are cyclical states that eventually resolve and identifying the precise moment you have left them is exactly what democratic governments are institutionally least equipped to do in real time — as the post-2020 inflation episode demonstrated with considerable force and a $1.9 trillion price tag.

The honest verdict

MMT correctly identifies that currency-issuing sovereigns face no nominal solvency constraint, and that private deleveraging requires a public deficit counterpart. These are genuine contributions. The politicians who ignore them deserve the recessions they engineer.

But then, from basic recognition of the mechanics and accounting, it goes on to build an entire policy superstructure on what is essentially accounting. This is rather akin to explaining double-entry bookkeeping as a concept then trying to reverse engineer how a sophisticated bank should look and be run. Further, the use of an inflation constraint is a serious flaw, especially one that is blind to the financial cycle dynamics, ignorant of the private credit dynamics, and dismissive of the intertemporal price signal, which is the mechanism through which capitalism allocates capital across time.

Many of its proponents also try to justify a framework that works in certain circumstances, such as a surplus creditor or in a liquidity trap, but it is not a general theory of macroeconomic management. It is a description of a narrow set of conditions with conclusions that are valid inside them and potentially catastrophic outside.

The printing press is always available. The question is whether those operating it understand what they are destroying in the process. The map, as ever, is not the territory. MMT has drawn a very fine map. It has simply mapped the wrong country, or it is just trying to exist in the wrong universe.

End notes:

[i] Even a perfectly measured inflation signal would still be monitoring the wrong variable. Claudio Borio's "The Financial Cycle and Macroeconomics" demonstrates that systemic instability announces itself through credit and property prices, not consumer prices — the financial cycle running on a 15–20 year frequency that inflation-targeting frameworks are structurally blind to. The empirical case is 1993–2007: every standard metric green, consumer prices contained, crisis among the worst in a century. Borio and Disyatat locate the cause not in Bernanke's savings glut but in the excess elasticity of the international monetary system — its inability to prevent imbalances from accumulating unchecked. MMT would institutionalise that elasticity as a design principle. "Monetary Policy in the Grip of a Pincer Movement" is the operational corollary: a framework that waits for inflation to validate tightening will be wrong in the worst direction at the worst times.

[ii] This also hits at the heart of MMT’s job creation scheme it creates artificial employment rather than productive employment - the distinction matters enormously and MMT largely ignores it. Employing people is not the same as employing people usefully, and a buffer stock of workers doing things the market would not willingly pay for is not a productive resource being held in reserve. It is a fiscal cost masquerading as a labour market policy. The second is that the theory is entirely silent on what work actually gets done. Finally, the job guarantee's structural problem is political as much as economic. Releasing buffer-stock workers when conditions warrant would require a government to preside over what every newspaper would call mass redundancies. It would not happen. The buffer becomes permanent, the fiscal cost structural, and the theoretical elegance delivers precisely the open-ended spending commitment MMT's critics always suspected was the destination.

[iii] Richard Vague's two centuries of crisis data across forty-plus countries establish the reliable predictor: not public debt ratios but rapid private credit growth — approximately 18 percentage points of GDP over five years. Fiscal expansion at low administered rates inflates assets; inflated assets become collateral; private leverage expands well beyond the initial impulse. An economy can satisfy every MMT condition while a credit cycle builds silently toward the outcome Vague's data consistently predicts.

Edward Chancellor's The Price of Time supplies the historical perspective: interest rate suppression is an ancient pathology with an invariant sequence — misallocation, asset inflation, wealth concentration, violent correction. John Law. The 1920s Fed. Japan. Dot-com. Subprime. The form changes; the dynamic does not. Drawing on Böhm-Bawerk, Chancellor shows investments made at a falsified cost of capital are not merely risky — they are predicated on conditions that never existed. When rates normalise, they are not repriced. They are revealed. MMT's position that the natural rate is simply whatever the central bank sets is not a theoretical refinement. It is a proposal to permanently administer away the economy's most important coordinating price.