The Fantasy of Modern Monetary Theory

Accounting Identities, Idle Resources, and the Professional Managerial Class

Modern Monetary Theory has attracted considerable attention in recent years as an intellectual framework that appears to offer unlimited fiscal freedom to sovereign governments. Its proponents claim that a government issuing its own fiat currency can never involuntarily default, that deficit spending is the source of private sector savings, and that the only real constraint on government expenditure is inflation, which can be managed through taxation. These claims rest on a series of errors: a confusion of accounting identities with causal economic laws, an inadequate theory of production, a naive model of idle resources, and a blindness to the distributional consequences of the policies it recommends. When examined carefully, MMT is less a coherent economic theory than an elaborate rationalization for the expansion of state power, one whose real-world application, most dramatically during the COVID-19 pandemic, has produced outcomes precisely opposite to those its advocates promised.

The Failure of Economic Prediction and the Problem of Dynamic Systems

Before engaging MMT directly, it is worth acknowledging the broader intellectual context in which it has flourished. Mainstream macroeconomics has never succeeded in building genuinely predictive models. Dynamic Stochastic General Equilibrium models, the dominant framework in academic and central bank economics for several decades, failed spectacularly to anticipate the 2008 financial crisis. Time series approaches perform adequately in-sample but notoriously poorly out-of-sample. Even sympathetic insiders such as Olivier Blanchard have acknowledged that DSGE models were constructed on deeply unrealistic microfoundations. The discipline has largely survived this predictive failure by retreating to the claim that it provides frameworks for thinking rather than genuine forecasts, a significant retreat from its scientific pretensions.

The deeper problem, however, is not merely that existing models are imperfect but that economics is the study of a high-dimensional, nonlinear, path-dependent dynamical system. Markets are not equilibrium phenomena amenable to comparative statics. They are processes unfolding through time, in which the decisions of agents alter the very parameters that other agents are trying to estimate. Prices are not just allocative signals but carriers of information across time. Investment decisions made today restructure the productive possibilities available tomorrow. None of this complexity is well-captured by the mathematical tools that mainstream economics has imported from physics, where parameters are stable and measurable.

MMT compounds this problem in a distinctive way. Its central analytical move is to treat a static national accounting identity as though it were a causal law governing a dynamic system. The sectoral balances equation, which states that the government deficit equals private sector net saving plus the current account surplus, is a true accounting identity in the same sense that a balance sheet must balance. It is true by definition. But MMT economists, most prominently Warren Mosler and Stephanie Kelton, use this identity as though it reveals something about how the economy works causally, as though cutting the government deficit must necessarily reduce private saving. This is like arguing about the direction of a river by staring at a photograph of it. The identity constrains outcomes but does not determine their causal structure. Robert Murphy, among others, has pointed out that the very same identity can be rearranged to support the traditional crowding-out argument, in which increased government borrowing reduces private investment. The identity itself is neutral between these interpretations. Choosing between them requires a theory, and MMT does not have an adequate one.

The Representative Agent, Homogeneous Capital, and the Invisibility of Malinvestment

Mainstream economics is guilty of its own theoretical failures that are directly relevant here. The representative agent framework, which models the entire economy as a single immortal household maximizing a utility function over an infinite horizon with access to a single homogeneous capital stock K, is not a simplification in any scientifically legitimate sense. A genuine simplification retains the essential structure of the phenomenon being studied while dropping irrelevant detail. The representative agent framework eliminates precisely the features that make macroeconomic phenomena interesting and real.

Malinvestment, the misallocation of capital into projects that appear profitable under distorted price signals but prove unsustainable when those distortions unwind, is inherently a phenomenon of heterogeneity. Different agents, with different local knowledge, different time horizons, and different access to credit, make different decisions simultaneously. When artificially low interest rates distort the intertemporal price of capital, they do not affect all investments uniformly. They systematically bias the structure of investment toward longer-horizon projects whose apparent profitability depends on the persistence of the distortion. This is a disaggregated, dynamic, multi-sector story. A model with a representative immortal household and homogeneous K cannot even represent it, let alone explain it.

The Austrian tradition, particularly through Hayek’s capital theory debates with Frank Knight, understood that capital is heterogeneous and time-structured. A factory built to produce steel is not costlessly convertible into a hospital. Malinvestment is painful because capital is specific: once misallocated, the loss is real and the reallocation is costly. The K in a DSGE model can be redeployed frictionlessly and instantaneously, which means the model has assumed away the entire problem of economic coordination that capital theory exists to analyze.

The Sonnenschein-Mantel-Debreu results, established in the 1970s, demonstrated that even granting all the assumptions of general equilibrium theory, individual demand curves cannot be aggregated into a well-behaved market demand curve without the representative agent assumption itself. The profession knew for fifty years that its foundations were shaky and largely carried on regardless, driven by the tractability of the framework and the career incentives of a discipline organized around mathematical publishability.

The Austrian Insight and Its Limits

The Austrian tradition offers the most serious intellectual challenge to both mainstream macroeconomics and MMT, but it is important to be precise about both its strengths and its weaknesses. Hayek’s critique of scientism in The Counter-Revolution of Science was essentially a claim that importing the methods of physics into social science is a category error. Physical systems have stable, measurable parameters. In a market, the very act of agents forming expectations, observing prices, and adjusting behavior changes the system being observed. The parameters are endogenous to the process being modeled. This is what makes markets fundamentally different from even complex physical systems.

Mises went further with praxeology, the claim that economic laws can be derived through logical deduction from the axiom of human action alone, without empirical investigation. This is a more problematic claim. The circularity objection, pressed by Terence Hutchison in the 1940s, has never been adequately answered. If the axioms of praxeology are definitional rather than empirical, the theorems derived from them are definitional as well. When Mises claims to derive time preference from the structure of action, he is unpacking a concept that was already contained in his definition, not discovering an independent fact about the world. The result is an elaborate tautology, which places praxeology in the same epistemological category as the MMT accounting identities it rightly criticizes. Murphy himself has reservations about praxeology on related grounds.

The most defensible Austrian position, stripped of praxeological ambition, is a set of well-grounded qualitative insights about how complex decentralized systems behave: that prices aggregate dispersed information that no central authority can possess, that credit expansion distorts the intertemporal structure of production in ways that eventually require painful correction, that unintended consequences systematically frustrate interventionist policies. These insights are probably correct in their broad outlines, and they are largely invisible to both MMT and mainstream DSGE economics. But they are heuristics, not a deductive science, and they resist the kind of formalization that would make them precisely testable.

The COVID Refutation: Idle Resources That Were Never Idle

MMT’s practical recommendations rest on the concept of idle resources. When an economy is operating below full employment, the argument goes, government spending or money creation can activate dormant productive capacity without generating inflation, because the spending draws on resources that are not otherwise being used. The COVID-19 pandemic constituted a natural experiment of unusual clarity, and its results were devastating for this framework.

When COVID struck, certain sectors collapsed suddenly: hospitality, travel, retail, live entertainment. Simultaneously, demand surged in other sectors: medical equipment, home goods, logistics infrastructure, semiconductor-dependent consumer electronics. The policy response, following Keynesian and MMT logic, was to pump enormous quantities of money into the system through direct transfers, expanded unemployment benefits, forgivable business loans, and central bank asset purchases. But the workers laid off from restaurants could not become semiconductor fabricators or logistics engineers overnight. The capital equipment in shuttered hotels could not be converted into hospital capacity. The result was not the activation of idle resources but a savage mismatch between where monetary demand was directed and where productive capacity actually existed. The inflation that followed was the predictable consequence.

MMT proponents had claimed that inflation was the only real constraint on government spending, and that this constraint was readily manageable through taxation. The inflation of 2021 and 2022 refuted both claims simultaneously. It proved rapid, broad-based, and extremely difficult to reverse without significant economic pain. The Federal Reserve’s belated and aggressive rate increases caused precisely the kind of disruption to investment and capital allocation that Austrian analysis would have anticipated. The idea that a government authority could fine-tune inflation in real time, identifying exactly when to withdraw money through taxation, was revealed as a technocratic fantasy with no basis in the actual dynamics of a complex monetary economy.

But the deeper problem runs beneath the inflation story. MMT, like Keynesianism more broadly, treats labor and capital as automatically productive once employed. This is a hidden assumption of enormous importance that is never examined. A worker paid to perform a task that generates no real value is employed in the accounting sense but contributes nothing to aggregate output. Much of the COVID fiscal response was closer to this description than its architects were willing to acknowledge. The Paycheck Protection Program kept businesses alive that had no viable future even before the pandemic. Stimulus transfers directed spending into supply-constrained sectors, bidding up prices rather than expanding real productive capacity. The assumption that government-directed monetary expansion would reliably find and activate genuinely productive idle resources was never theorized. It was assumed, and the assumption was wrong.

The question of marginal productivity is fundamental and systematically ignored in MMT’s framework. Whether a worker or unit of capital can be put to positive marginal productivity depends on whether there are complementary inputs available, whether the institutional environment supports productive coordination, whether the price signals guiding deployment are accurate, and whether the project being funded reflects genuine rather than artificially stimulated demand. None of these conditions is automatically satisfied by ensuring that aggregate demand is sufficient. Production requires not just spending but the right combination of labor, capital, knowledge, and institutional coordination, assembled in the right proportions at the right time. MMT has no theory of any of this.

The Professional Managerial Class and the Political Economy of MMT

The intellectual failures of MMT cannot be fully understood without attending to the social context in which it has flourished. Christopher Lasch’s analysis of the Professional Managerial Class is remarkably illuminating here. Lasch argued that the PMC, comprising credentialed professionals, policy experts, academics, NGO administrators, and bureaucrats, possesses a class interest in the expansion of state capacity that is largely invisible to itself because it is expressed in the language of altruism, expertise, and social justice. The PMC genuinely believes in what it advocates. But a class can have material interests that systematically shape its worldview without those interests being consciously acknowledged.

MMT is nearly a perfect expression of PMC ideology. It provides sophisticated-sounding intellectual justification for unlimited government spending, to be administered by experts who possess superior knowledge compared to the undisciplined signals of markets. It wraps these recommendations in progressive rhetoric about full employment, public investment, and social provision. The people who benefit most immediately from this framework are the PMC themselves: expanded bureaucracies require more administrators, more policy positions, more consultants, more academic grants. The very apparatus of expert management that MMT recommends is the apparatus that employs and empowers the PMC.

The distributional reality of the policies this ideology has supported over several decades is almost perfectly inverse to the stated intentions. Quantitative easing and near-zero interest rates, justified in the language of economic stimulus and recovery, massively inflated asset prices across every category: real estate, equities, private equity, collectibles, art. Households with significant asset holdings became dramatically wealthier. Households with no assets gained nothing from asset price appreciation. The monetary transmission mechanism that central bankers celebrated as accommodative was in practice a systematic transfer of wealth upward through the financial system.

Cheap debt compounded this effect. Wealthy individuals and corporations could borrow at near-zero rates and deploy capital into real assets or leveraged acquisitions, earning returns that vastly exceeded the cost of borrowing. Ordinary households, where they could access credit at all, paid much higher rates for consumer debt. The financial system transferred wealth upward through the interest rate structure while PMC economists applauded the policy as progressive.

Inflation then completed the regressive cycle. Wealthy asset holders were largely protected because their wealth was stored in real assets that appreciated with inflation. Wage earners and savers, particularly the working class and lower middle class who hold their wealth primarily in cash savings or wage income, saw their real purchasing power eroded. The people MMT claimed to champion suffered most from the inflation its prescriptions helped produce.

The concentration of ownership that has resulted from several decades of this policy regime is not an accident. Every expansion of state regulatory and fiscal capacity creates new surfaces for capture by the large corporations and financial institutions that possess the legal, political, and organizational resources to exploit them. The regulatory state does not restrain concentrated financial power; it creates barriers to entry that protect existing concentrations from competition. The tax code does not redistribute wealth; it provides sophisticated vehicles for wealth preservation that only the wealthy can access. MMT-style fiscal expansion does not empower workers; it channels money through institutions that well-connected actors are far better positioned to exploit than ordinary citizens.

Lasch’s tragedy was that he identified this pattern from the left, out of genuine concern for the working class and for democratic participation. His critique was not that inequality is acceptable but that the PMC had captured the language and institutions of progressive politics and redirected them toward its own class interests, leaving ordinary people without authentic representation while being constantly assured that expert advocates were acting on their behalf. The working class intuited this capture before the intellectuals acknowledged it, which is part of what drives the political realignments visible across Western democracies. MMT, with its promise of unlimited government spending administered by enlightened technocrats, crystallizes everything Lasch warned about.

Conclusion

Modern Monetary Theory fails on multiple levels simultaneously. At the level of economic methodology, it mistakes a static accounting identity for a causal law in a dynamic system, committing the same error it shares with the MMT critics who use the same identity to argue for crowding out. At the level of economic theory, it has no model of production adequate to the complexity of actual markets, no account of malinvestment, no theory of marginal productivity, and no understanding of how price signals coordinate dispersed knowledge. At the empirical level, the COVID-19 pandemic exposed the idle resources assumption as a fantasy, producing inflation where MMT predicted none was possible and misallocation where MMT promised activation of dormant capacity. At the political economy level, the policies MMT recommends have systematically served the interests of wealthy asset holders and credentialed administrators while harming the working-class constituencies in whose name they were proposed.

The honest position, uncomfortable as it is, may be that macroeconomics as a predictive science remains an aspiration rather than an achievement. The economy is a complex, nonlinear, path-dependent dynamical system that resists the modeling strategies that have dominated the discipline. The Austrian tradition has genuine qualitative insights about information, coordination, and the consequences of distortion, but has not succeeded in formalizing those insights into a rigorous and testable framework. Mainstream DSGE models achieve mathematical elegance at the cost of ontological accuracy. And MMT achieves rhetorical accessibility at the cost of basic theoretical coherence.

What all parties might acknowledge, if intellectual honesty prevailed, is that the gap between the complexity of actual economies and the adequacy of our models of them remains very large, and that policy confidence inversely proportional to that gap is a reliable indicator of ideology rather than science.

[embed]