Full Text
MONEY IS NOT THE ECONOMY: MEASURING NATIONAL SUCCESS BY PHYSICAL PRODUCTION, NOT FINANCIAL FLOWS
Many governments treat gross domestic product as the economy, although gross domestic product counts money flows rather than the production of goods, food, and energy that citizens need. Our paper asks: should governments target money-flow growth or productive-layer growth, meaning measurable output of goods, food, energy, housing, and healthcare? Our paper answers with the principle Money Is Not the Economy: money is infrastructure that helps the productive layer coordinate, and a society must never mistake the infrastructure for the destination. Aristotle separated household provisioning from money accumulation; Goodhart's law warns that measurements turned into targets stop measuring; research from the Bank for International Settlements shows that oversized financial sectors slow real growth. Our paper proposes two-layer national accounting: statistical offices reclassify instrumental-layer activity, meaning finance and administration, as intermediate cost, and measure productive-layer output in physical units such as homes completed, energy generated, and patients cured; existing wellbeing indices adjust money totals, while the proposed method counts real units. Open parameters include the weighting of intelligence production and culture production; the evaluation procedure extends material flow accounting, reclassifying published national statistics and comparing production-optimal policies against money-optimal policies. Our paper anticipates that national rankings change when financial churn stops counting as output, that ministries measured by physical results treat spending as cost rather than achievement, and that production-focused countries survive supply crises that money cannot solve. Our paper recommends productive-layer scoreboards, because expanding the supply of goods, food, energy, and housing makes citizens richer and returns national policy to the purpose of a country: citizens.
1. INTRODUCTION: THE CONFUSION BETWEEN MONEY AND THE ECONOMY
Modern governments steer nations by watching one number. The number is gross domestic product, and gross domestic product counts flows of money. When flows of money grow, ministers declare success; when flows of money shrink, ministers declare crisis. Yet citizens do not eat money, do not live inside money, and do not heat homes with money. Citizens eat food, live inside houses, and heat homes with energy. A government can preside over rising gross domestic product while citizens wait longer for houses, pay more for energy, and receive less care, because the number that the government watches measures the movement of money, not the production of the goods, the food, and the energy that citizens need.
Our paper asks one explicit research question: should national governments target the growth of money flows, or should national governments target the growth of the productive layer, meaning the measurable output of goods, food, energy, housing, and healthcare? Our paper argues for the productive layer, and our paper supplies a concrete measurement method, two-layer national accounting, so the argument becomes a policy instrument instead of a slogan.
Our paper makes four contributions. First, our paper formalizes the principle Money Is Not the Economy: money belongs to the instrumental layer of a national economy, production belongs to the productive layer, and the instrumental layer is justified exactly to the degree that the instrumental layer serves the productive layer. Second, our paper restores a lost classification argument: Simon Kuznets, the architect of national income accounting, originally treated government services and military spending as intermediate inputs rather than final output, and Kuznets lost the argument for political reasons, not conceptual reasons; our paper extends the intermediate-input logic of Kuznets to the entire instrumental layer. Third, our paper specifies the two-layer national accounting method: statistical offices reclassify instrumental-layer activity as intermediate cost and measure productive-layer output directly in physical units, building on the established practice of material flow accounting and physical input-output tables. Fourth, our paper derives policy implications: ministries measured by physical results treat spending as cost rather than achievement, supply expansion becomes the cheapest way to raise living standards, and production-focused countries hold the only insurance policy that pays out in every crisis.
The remainder of our paper proceeds as follows. Section 2 states the principle. Section 3 traces the intellectual lineage from Aristotle through Pigou, Tobin, and Kuznets. Section 4 explains why gross domestic product fails as a target. Section 5 presents evidence that money flows decouple from production. Section 6 presents evidence that production, not money, is the engine of national prosperity and national survival. Section 7 specifies the two-layer national accounting method. Section 8 states anticipated findings, the evaluation procedure, and named open parameters. Section 9 develops policy implications. Section 10 answers objections. Section 11 states limitations. Section 12 concludes.
2. THE PRINCIPLE: MONEY IS NOT THE ECONOMY
Our paper separates every national economy into two layers.
The productive layer contains the activities that create what human beings directly need and use: farms growing food, factories making goods, power plants generating energy, construction firms completing homes, hospitals curing patients, schools making students competent, laboratories producing knowledge, and artists producing culture. Within the productive layer, our paper places material provisioning first, because goods, food, and energy keep people alive; the production of intelligence and the production of culture stand on top of material provisioning and improve future material provisioning, but the production of intelligence and the production of culture cannot substitute for material provisioning.
The instrumental layer contains the activities that coordinate the productive layer without directly creating goods or services that people consume: money, banking, insurance, financial trading, taxation systems, compliance work, and administrative bureaucracy. Instrumental-layer activities are tools. A tool is justified exactly to the degree that the tool serves production.
Three plain sentences carry the principle. Money cannot be eaten. Money cannot heat a home. Money cannot cure a disease. Money only helps people exchange the food, the heat, and the medicine that farms, power plants, and laboratories must first produce.
Our paper immediately states the anti-misreading clause, because the most predictable objection misreads the principle as hostility toward money. The principle Money Is Not the Economy does not claim that money is unimportant. The principle claims that money is infrastructure, and good infrastructure must be cheap, reliable, and thin, and a society must never mistake the infrastructure for the destination. A road matters enormously, and a nation should build excellent roads; but a nation that measured national success by counting traffic instead of counting deliveries would soon optimize for congestion. Money is the road; production is the delivery.
One further clause completes the principle: instrumental-layer tools are thin but never neutral. Which credit exists determines what gets produced; how a marketplace is designed determines what gets sold; how a tax is collected determines who carries burden. Thinness therefore does not mean neglect. The thin layer must be designed extremely well, and the thin layer must also be measured honestly, as overhead, which is the accounting task of Section 7. For the individual citizen, one additional fact makes careful design a matter of justice: money stores life energy, because the citizen exchanged hours of work for money, and a companion paper in our research program (on the doctrine of Fiscal Secularity) develops the protections that stored life energy deserves.
3. INTELLECTUAL LINEAGE: ARISTOTLE, PIGOU, TOBIN, KUZNETS
The distinction between production and money is not new; the distinction is ancient, and the distinction was stated by the most authoritative voices in the history of economic thought.
Aristotle. In the introductory chapter of the Politics, Aristotle distinguished between oikonomía, the art of household management, and chrematistics, the art of acquisition; Karl Polanyi called the distinction "probably the most prophetic pointer ever made in the realm of the social sciences". Aristotle further separated two kinds of chrematistics: one acquisitive art that procures what is necessary for life, and another, lucrative chrematistics aimed at exchange and profit. Provisioning is bounded by need; money-making for money's own sake is unbounded, and Aristotle judged the unbounded form unnatural. The two-layer separation of our paper is the Aristotelian separation restated in the vocabulary of national accounting.
Pigou and the classical veil. Classical economists held that money acts merely as a medium of exchange and does not affect output and employment, the flow of goods and services, in any way; money was described as "a kind of wrapper, or garment in which goods came to us, a veil behind which the working of the real economic forces was concealed". Arthur Cecil Pigou gave the doctrine a book-length treatment in The Veil of Money, published by Macmillan in 1949, reviewed the same year in The Economic Journal, Volume 59, Issue 236, December 1949. Our paper does not adopt the strong classical claim that money never affects production, the housing evidence in Section 5 contradicts the strong claim, but our paper adopts the direction of the metaphor: the goods are the substance, and the money is the covering.
Tobin and functional efficiency. James Tobin delivered the modern statement in "On the efficiency of the financial system," Lloyds Bank Review 153, July 1984, pages 1–15. Tobin asked "whether the US and other high-income economies are devoting more resources to finance than can be justified by the gain in social welfare", and Tobin documented the pull of talent away from production: forty per cent of 1983 graduates of Yale's school of management took jobs in finance. Tobin's question, does the instrumental layer earn the resources the instrumental layer consumes, is the question that two-layer national accounting is built to answer continuously.
Kuznets, the warning, and the lost classification battle. Simon Kuznets designed national income accounting and immediately warned against the misuse of national income accounting. In the founding report, Kuznets, Simon, National Income, 1929–32, Senate Document No. 124, 73rd Congress, 2nd Session, U.S. Government Printing Office, Kuznets wrote: "The welfare of a nation can scarcely be inferred from a measurement of national income," and Kuznets listed what the measurement omitted: the distribution of income, the quality of goods produced, and the social value of the activities counted. Less famous, and central to our paper, is the classification practice of Kuznets. Kuznets excluded the state from national income calculations because Kuznets believed public services were an input for private production, since businesses factored the taxes paid for public services into product prices, inclusion would cause double counting, and the emerging Keynesian economists argued against Kuznets, because deficit-spending policy needed a measure of the state's contribution to economic activity. Kuznets likewise held that most peacetime military spending should be excluded from national income, because the military is an intermediate good rather than a final good like wheat or haircuts, the cost of the scarecrow that keeps the crows from the wheat is already included in the price of the wheat.
Kuznets lost the classification battle for reasons of policy convenience, not for reasons of accounting logic. Our paper reopens the battle and extends the intermediate-input logic of Kuznets from government services and military spending to the entire instrumental layer: finance, compliance, and administration are scarecrows, and a nation should count wheat.
4. WHY GROSS DOMESTIC PRODUCT FAILS AS A TARGET
Gross domestic product measures money flows, not welfare. Gross domestic product is useful for measuring marketed economic activity, but gross domestic product has been mistakenly used as a broader measure of welfare, which gross domestic product was never designed to measure. Gross domestic product measures mainly market transactions and ignores social costs, environmental impacts and income inequality. The inadequacy is officially recognized: the Commission on the Measurement of Economic Performance and Social Progress, created by the French Government in 2008 and led by Stiglitz, Sen and Fitoussi, examined how the wealth and social progress of a nation could be measured without relying on the uni-dimensional gross domestic product measure, and the Commission concluded that governments should move away from over-reliance on gross domestic product when assessing a country's health, towards a broader dashboard of indicators. The scientific flagship venues carry the same message: a comment in Nature argued that gross domestic product is a misleading measure of national success and that countries should act now to embrace new metrics.
A measurement turned into a target stops measuring. In the original 1975 form, Goodhart's law stated that any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes; the popular phrasing, due to Marilyn Strathern, runs "When a measure becomes a target, it ceases to be a good measure". The primary source is Goodhart, C.A.E. (1975), "Problems of Monetary Management: The UK Experience," in Papers in Monetary Economics, Volume I, Reserve Bank of Australia, reprinted in Monetary Theory and Practice, London: Macmillan, 1984, pages 91–121. The social-science precedent is Campbell, "Assessing the impact of planned social change," Evaluation and Program Planning, 1979, and a modern formal taxonomy is Manheim and Garrabrant, "Categorizing Variants of Goodhart's Law," 2018. Applied to national policy, the law reads: once leaders are rewarded for gross domestic product, leaders learn to grow the number rather than the substance behind the number. A country can raise gross domestic product through financial churn, for example, through houses trading back and forth at rising prices, while the number of houses, the supply of energy, and the supply of food stay unchanged. A thermometer taped to a radiator reports a warm house while the family freezes.
The paradox of persistence. Why does the number survive the criticism? Despite all theoretically and empirically motivated criticism of gross domestic product as a social welfare and progress indicator, the role of gross domestic product in economics, public policy, politics and society continues to be influential; most economists in business and government, teachers, journalists and policy makers of all political preferences continue to give much importance to gross domestic product and to call for unconditional growth of gross domestic product. Van den Bergh names the situation an information failure, and argues that removal of the information failure which gross domestic product represents, in monitoring economic progress and guiding public policy, will lead to decisions and developments more in line with improving human well-being. Our paper adopts the diagnosis: the problem is not that governments are wicked; the problem is that governments are steering with a corrupted instrument, and the corrupted instrument must be replaced at the level of official statistics, not at the level of speeches.
5. EVIDENCE THAT MONEY FLOWS DECOUPLE FROM PRODUCTION
If money flows and production always moved together, targeting either one would steer the nation equally well, and our paper would be unnecessary. The evidence shows that money flows and production do not always move together, and that beyond a threshold the two move apart.
Financialization is measured, not merely alleged. Systematic empirical evidence documents the financialization of the United States economy in the post-1970s period, where financialization is defined as a phase in capitalist development in which profits increasingly come from financial channels, with over forty percent of American economic profits being financial by the end of the millennium, reference: Krippner, G. R. (2005), "The Financialization of the American Economy," Socio-Economic Review, 3, 173–208. Profit, the money signal, detached from production, the physical substance.
Beyond a threshold, more finance means less growth. Research examining whether a threshold exists above which financial development no longer has a positive effect on economic growth finds that finance starts having a negative effect on output growth when credit to the private sector reaches one hundred percent of gross domestic product, published as Arcand, Berkes and Panizza (2015), "Too Much Finance?", Journal of Economic Growth, 20, 105–148. The Bank for International Settlements reaches the same conclusion twice: the level of financial development is good only up to a point, after which financial development becomes a drag on growth, and a fast-growing financial sector is detrimental to aggregate productivity growth; and, using a panel of twenty countries over thirty years, the faster the financial sector expands, the slower the real economy grows, references: Cecchetti and Kharroubi (2012), Reassessing the Impact of Finance on Growth, BIS Working Paper No. 381 and Cecchetti and Kharroubi (2015), "Why does financial sector growth crowd out real economic growth?", BIS Working Papers 490.
The magnitudes are official. Trading in over-the-counter foreign exchange markets reached 9.6 trillion United States dollars per day in April 2025, up twenty-eight percent from 7.5 trillion three years earlier, and global average daily turnover in interest rate derivatives reached 7.9 trillion United States dollars in April 2025. Combining the two instrument classes yields roughly 17.5 trillion dollars of trading per day; since world annual output is on the order of one hundred and ten trillion dollars, roughly one week of trading in two instrument classes matches one full year of world production (the comparison is our own arithmetic on the official turnover figures). A single-country illustration: in Chile, activity in the interest rate derivatives market reached a volume equivalent to 4.8 times the gross domestic product of the country. Our paper does not claim that every traded dollar is waste, hedging serves production, but a layer that turns over the world's annual output every few days cannot plausibly be, in aggregate, a proportional servant of production, and Tobin's functional-efficiency question returns with force.
Housing shows the decoupling inside household life. The housing affordability and wealth inequality crises facing advanced economies are driven by a feedback cycle between finance and landed property, created by the policy preference for private home ownership coupled with liberalization of bank credit, under which landed property becomes both the most attractive collateral for the banking system and the most desirable financial asset for households and investors. Building on Minsky's financial fragility hypothesis, land and house prices can rise without any active production, which dampens wider economic growth. Rising house prices raise gross domestic product statistics and household balance sheets while the physical stock of homes stands still, the exact confusion our paper names, played out in the largest purchase of every family's life.
6. EVIDENCE THAT PRODUCTION, NOT MONEY, IS THE ENGINE
Productive capabilities predict national income; money flows do not. The economic complexity research program provides the strongest quantitative support for production-first policy. Interpreting trade data as a network connecting countries to the products the countries export, measures of economic complexity are correlated with a country's level of income; deviations from the relationship are predictive of future growth, so countries tend to converge to the level of income dictated by the complexity of the productive structures of the countries, indicating that development efforts should focus on generating the conditions that allow complexity to emerge. The finding is robust: the economic complexity measure significantly explains current per-capita gross domestic product even with detailed controls for population, life expectancy, education and economy size, and highly predicts future long-term growth. Reference: Hidalgo and Hausmann (2009), "The Building Blocks of Economic Complexity," Proceedings of the National Academy of Sciences, 106(26), 10570–10575. In plain words: what a country can physically make predicts what a country will earn; production is the engine, and money is the reading on the dial.
Energy is a driver of growth, not a detail. Exergy analysis shows that useful work supplied to an economy is Granger-causal to output growth, and the associated research argues that much of twentieth-century economic growth, typically credited to general technological progress in models like Solow's, can be more accurately attributed to tangible improvements in converting primary energy into usable physical work. References: Warr and Ayres (2012), "Useful work and information as drivers of economic growth," Ecological Economics 73, 93–102; Warr, Ayres, Eisenmenger, Krausmann and Schandl (2010), "Energy use and economic development," Ecological Economics 69(10), 1904–1917; Ayres, van den Bergh, Lindenberger and Warr (2013), "The underestimated contribution of energy to economic growth," Structural Change and Economic Dynamics, 79–88. The goods-food-energy triad of Section 2 is therefore not rhetorical; energy sits inside the growth engine itself.
In a crisis, money cannot buy goods that nobody produces. The 2020 pandemic ran the experiment. Eighty countries and customs territories banned or limited the export of face masks, protective gear, gloves and other goods to mitigate shortages, as the World Trade Organization reported. Since the pandemic impacted virtually every country worldwide, there were no unaffected reservoirs of industrial activity to draw from. Dependence was extreme: China accounted for more than eighty-five percent of all United States imports of personal protective equipment and decreased exports, removing critical supplies from markets across the globe; reduced production and exports of pharmaceuticals and personal protective equipment from China exacerbated shortages of critical medical supplies. Money-rich, production-poor countries bid for goods that did not exist.
The 2021 semiconductor shortage repeated the experiment in a different sector. The semiconductor shortage cost the worldwide automotive industry an estimated two hundred and ten billion dollars in lost revenues in 2021, with production of 7.7 million vehicles lost; another estimate holds that more than 9.5 million units of global light-vehicle production were lost in 2021 as a direct result of a lack of necessary semiconductors. Crucially, money was present and goods were not: demand was strong, and together with accelerating prices and sharply declining inventories, the evidence pointed to an insufficient supply of new cars. Governments learned the lesson in the production vocabulary, not the money vocabulary: the United States government offered fifty billion dollars in incentives to Intel, Samsung, and the Taiwan Semiconductor Manufacturing Company to set up factories in the United States. Production is the only insurance policy that pays out in every crisis; money is an insurance policy that pays out only in good weather.
7. THE PROPOSAL: TWO-LAYER NATIONAL ACCOUNTING
Our paper proposes that national statistical offices maintain, alongside the existing monetary accounts, a second official account with two rules.
Rule one, reclassification. Instrumental-layer activity is booked as intermediate cost, never as final output. The instrumental layer is defined operationally: an activity belongs to the instrumental layer when the output of the activity is the coordination of other activities rather than a good or service that a person finally consumes. Finance, insurance, financial trading, tax compliance, and administrative overhead belong to the instrumental layer under the operational definition. The reclassification follows the original logic of Kuznets, Section 3 documented that Kuznets excluded the state from national income because public services were an input for private production whose inclusion would cause double counting, and extends the logic from government services to the entire coordination apparatus of the economy. Under rule one, financial churn can no longer appear as growth, for the same reason that the flour a bakery consumes cannot appear as bread.
Rule two, physical measurement. Productive-layer output is measured directly in physical and near-physical units. Headline indicators include: homes completed and dwelling floor area added; megawatt-hours of electricity generated, with capacity and import dependence stated; tonnes and calories of food produced; potable water delivered; passenger-kilometres and tonne-kilometres of transport provided; semiconductor units fabricated; hospital treatments completed and patients cured; students brought to certified competence. The units are chosen so that each headline answers a citizen-level question: can a family find a home, afford heat, eat well, travel, and be cured?
The method stands on existing statistical practice. Physical accounting is not an invention of our paper; physical accounting is an established branch of official statistics that our paper repurposes from environmental monitoring to the national scoreboard. Economy-wide material flow accounting is an established method to monitor resource use across scales, and the headline indicators of the method are widely used in policy; the framework provides consistent compilations of the material inputs to national economies, changes in material stocks within the economic system, and material outputs, reference: Krausmann, Schandl, Eisenmenger, Giljum and Jackson, "Material Flow Accounting: Measuring Global Material Use for Sustainable Development," Annual Review of Environment and Resources, volume 42 (2017). Deeper still, the physical input-output table records all physical flows associated with the economic activities defined in the System of National Accounts, flows of physical products, extraction, wastes and stock changes, and uses the same classification schemes as the monetary input-output table, so data from the physical table and the monetary table can be combined for analysis and policy modeling. National implementations already exist: comprehensive national input-output tables in purely physical units have been compiled for Germany, Finland and Denmark, beginning with the physical input-output tables of the German Federal Statistical Office (Stahmer, Kuhn and Braun, 1997), and the United Nations handbook on supply and use tables contains a chapter on compiling physical supply and use tables and environmentally extended input-output tables. Statistical offices therefore already possess the table structures; our paper asks statistical offices to promote the physical tables from environmental annex to national scoreboard, and to add the reclassification of rule one.
Measuring intelligence production. The production of intelligence, education, research, engineering skill, resists simple physical units, and our paper treats the weighting of intelligence production as a named open parameter rather than a solved problem. An established accounting literature nonetheless exists: the Jorgenson-Fraumeni approach measures human capital as the present discounted value of the lifetime earnings of all living people in an economy, proposing the increment to lifetime labor incomes as the measure of the output of education and training, and later work integrates Jorgenson-Fraumeni human capital into a complete national income accounting system, revealing a fresh perspective on economic growth across time. Two-layer national accounting can adopt the lifetime-income method as the initial weighting for intelligence production while flagging the money-denominated character of the method as a compromise awaiting better physical proxies (certified competences, completed apprenticeships, peer-reviewed findings).
Differentiation from existing alternatives. Alternatives to gross domestic product exist and are respected: governments and organizations already follow the Human Development Index, the Better Life Index and the Genuine Progress Indicator. The Genuine Progress Indicator, the most economically ambitious alternative, starts with the same personal consumption data as gross domestic product but adjusts for income distribution, adds the value of household and volunteer work, and subtracts costs such as crime and pollution. Every named alternative is therefore a money-denominated welfare adjustment: the alternatives correct the price total. Two-layer national accounting makes a different move: two-layer national accounting refuses the price total as the headline, counts real units on the productive layer, and books the instrumental layer as overhead. The two moves are complementary, a country can publish a Genuine Progress Indicator and a two-layer account side by side, but the moves are not the same move, and the difference is the novelty claim of our paper.
What our paper does not change. Two-layer national accounting changes the scoreboard of the state; two-layer national accounting does not abolish market prices, does not replace private bookkeeping, and does not plan production. Firms keep trading at market prices; the monetary national accounts continue to be compiled; the physical account and the monetary account share classifications by design, as the physical input-output literature already demonstrates. Section 10 returns to the point when answering the central-planning objection.
8. ANTICIPATED FINDINGS, EVALUATION PROCEDURE, AND OPEN PARAMETERS
Our paper is a position paper with a defined empirical program. Three findings are anticipated.
Anticipated finding one, rankings shift. When instrumental-layer activity is reclassified as intermediate cost, countries with large financial centers fall relative to the money-denominated rankings of the same countries, and countries with strong physical output per person rise. The size of the shift measures how much of the recorded prosperity of each country is coordination overhead rather than production.
Anticipated finding two, ministry incentives flip. Where ministries are evaluated on physical results, homes completed, energy generated, patients cured, students competent, spending changes sign in the evaluation: spending becomes a cost to minimize rather than an achievement to report. Nobody praises a bakery for the flour the bakery consumed; under two-layer national accounting, no ministry is praised for the budget the ministry consumed.
Anticipated finding three, production predicts crisis survival. Countries ranking high on the productive-layer scoreboard before 2020 experienced smaller shortage shocks during the pandemic and the semiconductor crisis than countries ranking high on money-denominated scoreboards, because physical capacity, the thing the new scoreboard counts, is the thing that shortages test.
Evaluation procedure. The procedure has four steps and uses only published statistics, so any statistical office or research team can replicate the procedure. Step one: for a pilot set of five to ten countries, rebuild national accounts under rule one, using existing supply-and-use tables to identify instrumental-layer value added and reclassify the identified value added as intermediate consumption. Step two: assemble the physical headline indicators of rule two from existing material flow accounts, energy balances, construction statistics, agricultural statistics, and health statistics. Step three: compare country rankings and trend lines under the monetary account and under the two-layer account, and measure the divergence. Step four: run the two historical event studies, the 2020 medical supply shock and the 2021 semiconductor shock, and test whether the two-layer scoreboard, computed with pre-crisis data only, predicts crisis performance better than gross domestic product predicts crisis performance.
Named open parameters. First, the weighting of intelligence production and culture production (Section 7 states the candidate method and the reservation about the candidate method). Second, quality adjustment: a home completed in 2026 is not a home completed in 1970, and hedonic correction of physical units imports prices through the back door; the degree of permitted hedonic correction is a design dial, not a settled constant. Third, boundary assignments: some activities serve both layers (software, logistics, wholesale trade), and the classification manual must assign the dual activities by final-use tests, with sensitivity analysis published for contested assignments. Fourth, aggregation: whether the physical headlines remain a dashboard or are compressed into a small number of indices, and by which weights, our default position is the dashboard, consistent with the recommendation of the Stiglitz-Sen-Fitoussi Commission to move towards a broader dashboard of indicators.
9. IMPLICATIONS FOR GOVERNMENT POLICY
9.1 The metric flip inside ministries. The most direct application of two-layer national accounting is administrative: evaluate public organizations on productive-layer results. The evidence on public-sector targets is two-sided, and our paper uses both sides. The cautionary side: Bevan and Hood (2006), "What's Measured Is What Matters: Targets and Gaming in the English Public Health Care System," Public Administration, 84, 517–538 documents how target regimes invite gaming, summarized as improvements in reported performance accompanied by attempts to game the system, such as "hitting the target and missing the point". The constructive side: targets with sanctions can work when the targets accurately measure key dimensions of performance that cannot easily be manipulated by gaming, examples include hospital waiting times and ambulance response times, and in the English natural experiment by 2002 hospital waiting times had decreased in England but increased in Wales. The design lesson for two-layer national accounting is exact: choose final physical outcomes (a completed home, a cured patient) rather than clerical proxies, publish several indicators together rather than one, and audit on randomized schedules. Physical finality is the gaming resistance.
9.2 Aligning the treasury with production. A companion paper in our research program proposes replacing complex taxation with a flat levy collected automatically on each transaction through a central-bank payment system, at a rate consistent with historically persistent flat levies. Under the combined design, government revenue grows only when real transactions grow, so the treasury acquires a direct structural interest in expanding production and exchange rather than inflating asset prices. Two-layer national accounting supplies the scoreboard; the transaction levy supplies the incentive; the two instruments point the state at the same target, the productive layer.
9.3 Supply expansion is the silent raise. The cheapest way to make citizens richer is to make the goods citizens buy more plentiful and therefore cheaper, and housing is the leading case. Glaeser and Gyourko (2018), "The Economic Implications of Housing Supply," Journal of Economic Perspectives 32(1), 3–30 show that the gap between house price and production cost can be understood as a regulatory tax, and the implicit tax on development created by housing regulations is higher in many areas than any reasonable negative externalities associated with new construction. A government watching money flows sees rising house prices as wealth; a government watching the productive layer sees a regulatory tax on the young and a construction deficit to close. The same logic covers energy: megawatt-hours generated, not energy-sector revenue, is the number that predicts whether factories run and homes stay warm, and Section 6 established that useful energy is a driver of growth itself.
9.4 Health: buying cures instead of renting health. Money-flow metrics reward permanent treatment; production metrics reward cure. The finance literature states the distinction plainly: taking traditional medications is like renting an apartment, paying for the benefit a month at a time, while curative therapies are akin to buying a house, and the difference is between "renting health" one pill at a time and buying a cure; reference: Montazerhodjat, Weinstock and Lo (2016), on financing curative therapies, Science Translational Medicine. The market's own analysts confirmed the incentive problem: a 2018 Goldman Sachs biotechnology report asked, "Is curing patients a sustainable business model?", observing that hepatitis C treatments achieving cure rates above ninety percent saw United States sales peak at 12.5 billion dollars in 2015 and fall steadily thereafter because curing patients means no recurring revenue and, for infectious disease, shrinks the pool of new patients. Under two-layer national accounting the paradox dissolves: the health system is scored on patients cured, a productive-layer output, so public subsidy and public procurement flow toward cures, prevention, and rejuvenation, and away from permanent symptom subscriptions. Citizens get healthier; recorded money flows in the health sector may well shrink; under the new scoreboard, shrinkage of overhead while cures rise is registered as success, which is the whole point.
9.5 Attacking the ratchet of the state share. Two robust regularities explain why the instrumental share of national life ratchets upward. First, Wagner's law, the law of increasing state activity, observes that public expenditure increases as national income rises, with modern panel support in bootstrap Granger-causality results favoring Wagner's law in seven of twelve OECD countries, carrying the policy implication that aging populations and welfare demand may force higher taxes or excessive borrowing. Second, the Baumol effect: wages in jobs with little productivity growth rise in response to rising wages in progressive sectors, so services, health, education, care, administration, become relatively more expensive over time, a mechanism with strong professional consensus: when weighted by confidence, eighty-eight percent of a polled panel of United States economic experts agreed that rising manufacturing productivity leads the cost of labor-intensive services to rise. The overhead is not small: compliance with the United States federal tax code consumes 7.9 billion hours plus out-of-pocket costs, totaling roughly 546 billion dollars per year, nearly two percent of gross domestic product; Canadian estimates place private business compliance costs between 1.2 and 1.8 percent of gross domestic product, with public administration adding 0.2 to 0.5 percent; and the burden is regressive: as a proportion of taxes paid, compliance costs of medium-sized businesses are significantly higher than for the largest businesses (see also Slemrod, "Tax Compliance and Enforcement," Journal of Economic Literature 57(4), 904–954, 2019). Two-layer national accounting makes the ratchet visible year by year, the instrumental layer is a published overhead ratio, and administrative automation becomes the one form of automation with no demand-side downside, because bureaucracy has no consumers to lose: every bureaucrat-hour freed is pure gain for the productive layer.
9.6 Strategic capacity before the crisis, not after. Section 6 recorded that governments responded to the semiconductor shock with fifty billion dollars of factory incentives, production policy adopted in panic, after the shortage. A productive-layer scoreboard publishes chip fabrication, food self-sufficiency, energy capacity, and medical manufacturing continuously, so legislators see the vulnerability while repair is still cheap. Under money-flow accounting, resilience is invisible until the day resilience is priceless.
10. OBJECTIONS AND ANSWERS
Objection one: "Physical targets are Soviet planning." The objection deserves a direct answer. Soviet material planning replaced market prices, private property, and consumer choice with administrative production orders; the failure of Soviet material planning is documented, and the target-gaming pathologies of Soviet planning are the canonical illustrations of Goodhart's law, indeed the public-administration literature draws the parallel explicitly, noting that the English targets regime had obvious parallels with the Soviet regime, resting on synecdoche, taking a part to stand for the whole, and on the assumption that measurement problems and gaming do not matter. Two-layer national accounting shares no instrument with Soviet planning: markets keep setting prices, firms keep choosing outputs, consumers keep choosing purchases, and the monetary accounts keep being published. The only thing that changes is the scoreboard by which citizens judge governments and governments judge ministries. A nation that counts homes is not a nation that orders homes; a nation that counts homes is a nation that notices when homes are missing.
Objection two: "Goodhart's law will corrupt your physical metrics too." Correct as a warning, and Section 9.1 already contains the answer: the empirical record shows that gaming concentrates where measures are clerical proxies, while targets measuring key dimensions that cannot easily be manipulated, waiting times, response times, produced real improvement. Two-layer national accounting therefore selects final physical outcomes, publishes dashboards rather than single numbers, randomizes audit timing, and treats every metric as reviewable. Goodhart's law counsels careful metric design; Goodhart's law does not counsel steering by a corrupted metric because all metrics are imperfect.
Objection three: "Prices aggregate dispersed information; physical units discard the information." The Hayekian point about prices is accepted in full, for market coordination. Firms and households should keep using prices, and under our proposal firms and households do keep using prices. The claim of our paper is narrower: the aggregate of money flows is a corrupted summary statistic for national success, because Section 5 showed that money flows decouple from production at the aggregate level. Nothing in the price system requires that the state's headline scoreboard be a money total; the physical input-output literature demonstrates that physical tables and monetary tables share classification schemes and are combined for analysis, the two informations coexist by construction.
Objection four: "You cannot add homes to megawatt-hours to cured patients." Correct, and our paper does not add homes to megawatt-hours to cured patients. The productive-layer account is a dashboard of headline physical indicators, following the dashboard recommendation of the Stiglitz-Sen-Fitoussi Commission. Aggregation weights, where a compressed index is politically demanded, are a named open parameter with published sensitivity analysis. A dashboard that refuses false precision is more honest than a single number that hides the refusal.
Objection five: "Finance matters; your proposal starves finance." Our paper claims calibration, not abolition. The threshold evidence cuts both ways and our paper cites both ways: the level of financial development is good up to a point, finance below the threshold serves production, and credit that funds a factory passes the service test of Section 2. Two-layer national accounting simply prices the instrumental layer as what the instrumental layer is, a cost, so a nation can ask continuously whether each additional unit of coordination overhead buys additional production. Tobin asked the question once in 1984; the scoreboard asks the question every year.
Objection six: "Existing indices already fix gross domestic product." Section 7 answered the objection in detail: the existing indices are money-denominated welfare adjustments, while two-layer national accounting is a physical-unit scoreboard plus an intermediate-cost reclassification, reviving the original classification practice of Kuznets. Complementary, not identical.
Objection seven: "Physical units ignore quality and new goods." Partly correct, and stated as an open parameter: hedonic and quality adjustment of physical units is permitted by design dial, and the intelligence-production account leans on the lifetime-income measure of education output until better physical proxies mature. Our paper accepts a rough measure of the right thing over a precise measure of the wrong thing.
11. LIMITATIONS
Our paper carries five limitations, stated plainly. First, culture production and care work resist both money measures and physical measures; the Genuine Progress Indicator tradition of adding the value of household and volunteer work marks the gap without closing the gap, and two-layer national accounting inherits the gap. Second, the reclassification boundary of the instrumental layer requires a classification manual with contested cases, and classification manuals attract lobbying; the defense is publication of sensitivity analyses, not the pretense of a lobby-proof manual. Third, international comparability requires standardization work on the scale of the existing System of National Accounts revisions, a decade-class project, honestly stated. Fourth, the evaluation procedure of Section 8 tests predictive validity on two crises; two crises are suggestive, not conclusive. Fifth, the deepest obstacle is political rather than technical: policy makers and politicians of all political preferences continue to give much importance to gross domestic product, and a scoreboard that reclassifies large industries as overhead will be resisted by the industries so reclassified. Our paper's answer to the fifth limitation is the whole argument: the measurement corruption is precisely why the correction must live in official statistics rather than in discretionary politics.
12. CONCLUSION
Money is not the economy. The economy is farms, factories, power plants, homes, hospitals, and schools, the productive layer, and money is the instrumental layer that helps the productive layer coordinate. Gross domestic product, a money total, was introduced with a warning from the inventor of gross domestic product, "The welfare of a nation can scarcely be inferred from a measurement of national income", and ninety years of measurement corruption, financialization, and crisis evidence have proven the warning. Our paper proposes the correction the inventor himself began: reclassify coordination as cost, count production in real units, and let governments be judged by homes completed, energy generated, food produced, and patients cured. The instruments exist in the statistical offices already; the evidence base exists in the journals already; the missing element is the decision. A country that adopts the two-layer scoreboard does not become a planned economy; the country becomes a seeing economy, and a country that sees production clearly will choose production, because production, not money, is what keeps citizens alive, housed, warm, and well. The purpose of a country is the citizens of the country.
Many governments treat gross domestic product as the economy, although gross domestic product counts money flows rather than the production of goods, food, and energy that citizens need. Our paper asks: should governments target money-flow growth or productive-layer growth, meaning measurable output of goods, food, energy, housing, and healthcare? Our paper answers with the principle Money Is Not the Economy: money is infrastructure that helps the productive layer coordinate, and a society must never mistake the infrastructure for the destination. Aristotle separated household provisioning from money accumulation; Goodhart's law warns that measurements turned into targets stop measuring; research from the Bank for International Settlements shows that oversized financial sectors slow real growth. Our paper proposes two-layer national accounting: statistical offices reclassify instrumental-layer activity, meaning finance and administration, as intermediate cost, and measure productive-layer output in physical units such as homes completed, energy generated, and patients cured; existing wellbeing indices adjust money totals, while the proposed method counts real units. Open parameters include the weighting of intelligence production and culture production; the evaluation procedure extends material flow accounting, reclassifying published national statistics and comparing production-optimal policies against money-optimal policies. Our paper anticipates that national rankings change when financial churn stops counting as output, that ministries measured by physical results treat spending as cost rather than achievement, and that production-focused countries survive supply crises that money cannot solve. Our paper recommends productive-layer scoreboards, because expanding the supply of goods, food, energy, and housing makes citizens richer and returns national policy to the purpose of a country: citizens.
1. INTRODUCTION: THE CONFUSION BETWEEN MONEY AND THE ECONOMY
Modern governments steer nations by watching one number. The number is gross domestic product, and gross domestic product counts flows of money. When flows of money grow, ministers declare success; when flows of money shrink, ministers declare crisis. Yet citizens do not eat money, do not live inside money, and do not heat homes with money. Citizens eat food, live inside houses, and heat homes with energy. A government can preside over rising gross domestic product while citizens wait longer for houses, pay more for energy, and receive less care, because the number that the government watches measures the movement of money, not the production of the goods, the food, and the energy that citizens need.
Our paper asks one explicit research question: should national governments target the growth of money flows, or should national governments target the growth of the productive layer, meaning the measurable output of goods, food, energy, housing, and healthcare? Our paper argues for the productive layer, and our paper supplies a concrete measurement method, two-layer national accounting, so the argument becomes a policy instrument instead of a slogan.
Our paper makes four contributions. First, our paper formalizes the principle Money Is Not the Economy: money belongs to the instrumental layer of a national economy, production belongs to the productive layer, and the instrumental layer is justified exactly to the degree that the instrumental layer serves the productive layer. Second, our paper restores a lost classification argument: Simon Kuznets, the architect of national income accounting, originally treated government services and military spending as intermediate inputs rather than final output, and Kuznets lost the argument for political reasons, not conceptual reasons; our paper extends the intermediate-input logic of Kuznets to the entire instrumental layer. Third, our paper specifies the two-layer national accounting method: statistical offices reclassify instrumental-layer activity as intermediate cost and measure productive-layer output directly in physical units, building on the established practice of material flow accounting and physical input-output tables. Fourth, our paper derives policy implications: ministries measured by physical results treat spending as cost rather than achievement, supply expansion becomes the cheapest way to raise living standards, and production-focused countries hold the only insurance policy that pays out in every crisis.
The remainder of our paper proceeds as follows. Section 2 states the principle. Section 3 traces the intellectual lineage from Aristotle through Pigou, Tobin, and Kuznets. Section 4 explains why gross domestic product fails as a target. Section 5 presents evidence that money flows decouple from production. Section 6 presents evidence that production, not money, is the engine of national prosperity and national survival. Section 7 specifies the two-layer national accounting method. Section 8 states anticipated findings, the evaluation procedure, and named open parameters. Section 9 develops policy implications. Section 10 answers objections. Section 11 states limitations. Section 12 concludes.
2. THE PRINCIPLE: MONEY IS NOT THE ECONOMY
Our paper separates every national economy into two layers.
The productive layer contains the activities that create what human beings directly need and use: farms growing food, factories making goods, power plants generating energy, construction firms completing homes, hospitals curing patients, schools making students competent, laboratories producing knowledge, and artists producing culture. Within the productive layer, our paper places material provisioning first, because goods, food, and energy keep people alive; the production of intelligence and the production of culture stand on top of material provisioning and improve future material provisioning, but the production of intelligence and the production of culture cannot substitute for material provisioning.
The instrumental layer contains the activities that coordinate the productive layer without directly creating goods or services that people consume: money, banking, insurance, financial trading, taxation systems, compliance work, and administrative bureaucracy. Instrumental-layer activities are tools. A tool is justified exactly to the degree that the tool serves production.
Three plain sentences carry the principle. Money cannot be eaten. Money cannot heat a home. Money cannot cure a disease. Money only helps people exchange the food, the heat, and the medicine that farms, power plants, and laboratories must first produce.
Our paper immediately states the anti-misreading clause, because the most predictable objection misreads the principle as hostility toward money. The principle Money Is Not the Economy does not claim that money is unimportant. The principle claims that money is infrastructure, and good infrastructure must be cheap, reliable, and thin, and a society must never mistake the infrastructure for the destination. A road matters enormously, and a nation should build excellent roads; but a nation that measured national success by counting traffic instead of counting deliveries would soon optimize for congestion. Money is the road; production is the delivery.
One further clause completes the principle: instrumental-layer tools are thin but never neutral. Which credit exists determines what gets produced; how a marketplace is designed determines what gets sold; how a tax is collected determines who carries burden. Thinness therefore does not mean neglect. The thin layer must be designed extremely well, and the thin layer must also be measured honestly, as overhead, which is the accounting task of Section 7. For the individual citizen, one additional fact makes careful design a matter of justice: money stores life energy, because the citizen exchanged hours of work for money, and a companion paper in our research program (on the doctrine of Fiscal Secularity) develops the protections that stored life energy deserves.
3. INTELLECTUAL LINEAGE: ARISTOTLE, PIGOU, TOBIN, KUZNETS
The distinction between production and money is not new; the distinction is ancient, and the distinction was stated by the most authoritative voices in the history of economic thought.
Aristotle. In the introductory chapter of the Politics, Aristotle distinguished between oikonomía, the art of household management, and chrematistics, the art of acquisition; Karl Polanyi called the distinction "probably the most prophetic pointer ever made in the realm of the social sciences". Aristotle further separated two kinds of chrematistics: one acquisitive art that procures what is necessary for life, and another, lucrative chrematistics aimed at exchange and profit. Provisioning is bounded by need; money-making for money's own sake is unbounded, and Aristotle judged the unbounded form unnatural. The two-layer separation of our paper is the Aristotelian separation restated in the vocabulary of national accounting.
Pigou and the classical veil. Classical economists held that money acts merely as a medium of exchange and does not affect output and employment, the flow of goods and services, in any way; money was described as "a kind of wrapper, or garment in which goods came to us, a veil behind which the working of the real economic forces was concealed". Arthur Cecil Pigou gave the doctrine a book-length treatment in The Veil of Money, published by Macmillan in 1949, reviewed the same year in The Economic Journal, Volume 59, Issue 236, December 1949. Our paper does not adopt the strong classical claim that money never affects production, the housing evidence in Section 5 contradicts the strong claim, but our paper adopts the direction of the metaphor: the goods are the substance, and the money is the covering.
Tobin and functional efficiency. James Tobin delivered the modern statement in "On the efficiency of the financial system," Lloyds Bank Review 153, July 1984, pages 1–15. Tobin asked "whether the US and other high-income economies are devoting more resources to finance than can be justified by the gain in social welfare", and Tobin documented the pull of talent away from production: forty per cent of 1983 graduates of Yale's school of management took jobs in finance. Tobin's question, does the instrumental layer earn the resources the instrumental layer consumes, is the question that two-layer national accounting is built to answer continuously.
Kuznets, the warning, and the lost classification battle. Simon Kuznets designed national income accounting and immediately warned against the misuse of national income accounting. In the founding report, Kuznets, Simon, National Income, 1929–32, Senate Document No. 124, 73rd Congress, 2nd Session, U.S. Government Printing Office, Kuznets wrote: "The welfare of a nation can scarcely be inferred from a measurement of national income," and Kuznets listed what the measurement omitted: the distribution of income, the quality of goods produced, and the social value of the activities counted. Less famous, and central to our paper, is the classification practice of Kuznets. Kuznets excluded the state from national income calculations because Kuznets believed public services were an input for private production, since businesses factored the taxes paid for public services into product prices, inclusion would cause double counting, and the emerging Keynesian economists argued against Kuznets, because deficit-spending policy needed a measure of the state's contribution to economic activity. Kuznets likewise held that most peacetime military spending should be excluded from national income, because the military is an intermediate good rather than a final good like wheat or haircuts, the cost of the scarecrow that keeps the crows from the wheat is already included in the price of the wheat.
Kuznets lost the classification battle for reasons of policy convenience, not for reasons of accounting logic. Our paper reopens the battle and extends the intermediate-input logic of Kuznets from government services and military spending to the entire instrumental layer: finance, compliance, and administration are scarecrows, and a nation should count wheat.
4. WHY GROSS DOMESTIC PRODUCT FAILS AS A TARGET
Gross domestic product measures money flows, not welfare. Gross domestic product is useful for measuring marketed economic activity, but gross domestic product has been mistakenly used as a broader measure of welfare, which gross domestic product was never designed to measure. Gross domestic product measures mainly market transactions and ignores social costs, environmental impacts and income inequality. The inadequacy is officially recognized: the Commission on the Measurement of Economic Performance and Social Progress, created by the French Government in 2008 and led by Stiglitz, Sen and Fitoussi, examined how the wealth and social progress of a nation could be measured without relying on the uni-dimensional gross domestic product measure, and the Commission concluded that governments should move away from over-reliance on gross domestic product when assessing a country's health, towards a broader dashboard of indicators. The scientific flagship venues carry the same message: a comment in Nature argued that gross domestic product is a misleading measure of national success and that countries should act now to embrace new metrics.
A measurement turned into a target stops measuring. In the original 1975 form, Goodhart's law stated that any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes; the popular phrasing, due to Marilyn Strathern, runs "When a measure becomes a target, it ceases to be a good measure". The primary source is Goodhart, C.A.E. (1975), "Problems of Monetary Management: The UK Experience," in Papers in Monetary Economics, Volume I, Reserve Bank of Australia, reprinted in Monetary Theory and Practice, London: Macmillan, 1984, pages 91–121. The social-science precedent is Campbell, "Assessing the impact of planned social change," Evaluation and Program Planning, 1979, and a modern formal taxonomy is Manheim and Garrabrant, "Categorizing Variants of Goodhart's Law," 2018. Applied to national policy, the law reads: once leaders are rewarded for gross domestic product, leaders learn to grow the number rather than the substance behind the number. A country can raise gross domestic product through financial churn, for example, through houses trading back and forth at rising prices, while the number of houses, the supply of energy, and the supply of food stay unchanged. A thermometer taped to a radiator reports a warm house while the family freezes.
The paradox of persistence. Why does the number survive the criticism? Despite all theoretically and empirically motivated criticism of gross domestic product as a social welfare and progress indicator, the role of gross domestic product in economics, public policy, politics and society continues to be influential; most economists in business and government, teachers, journalists and policy makers of all political preferences continue to give much importance to gross domestic product and to call for unconditional growth of gross domestic product. Van den Bergh names the situation an information failure, and argues that removal of the information failure which gross domestic product represents, in monitoring economic progress and guiding public policy, will lead to decisions and developments more in line with improving human well-being. Our paper adopts the diagnosis: the problem is not that governments are wicked; the problem is that governments are steering with a corrupted instrument, and the corrupted instrument must be replaced at the level of official statistics, not at the level of speeches.
5. EVIDENCE THAT MONEY FLOWS DECOUPLE FROM PRODUCTION
If money flows and production always moved together, targeting either one would steer the nation equally well, and our paper would be unnecessary. The evidence shows that money flows and production do not always move together, and that beyond a threshold the two move apart.
Financialization is measured, not merely alleged. Systematic empirical evidence documents the financialization of the United States economy in the post-1970s period, where financialization is defined as a phase in capitalist development in which profits increasingly come from financial channels, with over forty percent of American economic profits being financial by the end of the millennium, reference: Krippner, G. R. (2005), "The Financialization of the American Economy," Socio-Economic Review, 3, 173–208. Profit, the money signal, detached from production, the physical substance.
Beyond a threshold, more finance means less growth. Research examining whether a threshold exists above which financial development no longer has a positive effect on economic growth finds that finance starts having a negative effect on output growth when credit to the private sector reaches one hundred percent of gross domestic product, published as Arcand, Berkes and Panizza (2015), "Too Much Finance?", Journal of Economic Growth, 20, 105–148. The Bank for International Settlements reaches the same conclusion twice: the level of financial development is good only up to a point, after which financial development becomes a drag on growth, and a fast-growing financial sector is detrimental to aggregate productivity growth; and, using a panel of twenty countries over thirty years, the faster the financial sector expands, the slower the real economy grows, references: Cecchetti and Kharroubi (2012), Reassessing the Impact of Finance on Growth, BIS Working Paper No. 381 and Cecchetti and Kharroubi (2015), "Why does financial sector growth crowd out real economic growth?", BIS Working Papers 490.
The magnitudes are official. Trading in over-the-counter foreign exchange markets reached 9.6 trillion United States dollars per day in April 2025, up twenty-eight percent from 7.5 trillion three years earlier, and global average daily turnover in interest rate derivatives reached 7.9 trillion United States dollars in April 2025. Combining the two instrument classes yields roughly 17.5 trillion dollars of trading per day; since world annual output is on the order of one hundred and ten trillion dollars, roughly one week of trading in two instrument classes matches one full year of world production (the comparison is our own arithmetic on the official turnover figures). A single-country illustration: in Chile, activity in the interest rate derivatives market reached a volume equivalent to 4.8 times the gross domestic product of the country. Our paper does not claim that every traded dollar is waste, hedging serves production, but a layer that turns over the world's annual output every few days cannot plausibly be, in aggregate, a proportional servant of production, and Tobin's functional-efficiency question returns with force.
Housing shows the decoupling inside household life. The housing affordability and wealth inequality crises facing advanced economies are driven by a feedback cycle between finance and landed property, created by the policy preference for private home ownership coupled with liberalization of bank credit, under which landed property becomes both the most attractive collateral for the banking system and the most desirable financial asset for households and investors. Building on Minsky's financial fragility hypothesis, land and house prices can rise without any active production, which dampens wider economic growth. Rising house prices raise gross domestic product statistics and household balance sheets while the physical stock of homes stands still, the exact confusion our paper names, played out in the largest purchase of every family's life.
6. EVIDENCE THAT PRODUCTION, NOT MONEY, IS THE ENGINE
Productive capabilities predict national income; money flows do not. The economic complexity research program provides the strongest quantitative support for production-first policy. Interpreting trade data as a network connecting countries to the products the countries export, measures of economic complexity are correlated with a country's level of income; deviations from the relationship are predictive of future growth, so countries tend to converge to the level of income dictated by the complexity of the productive structures of the countries, indicating that development efforts should focus on generating the conditions that allow complexity to emerge. The finding is robust: the economic complexity measure significantly explains current per-capita gross domestic product even with detailed controls for population, life expectancy, education and economy size, and highly predicts future long-term growth. Reference: Hidalgo and Hausmann (2009), "The Building Blocks of Economic Complexity," Proceedings of the National Academy of Sciences, 106(26), 10570–10575. In plain words: what a country can physically make predicts what a country will earn; production is the engine, and money is the reading on the dial.
Energy is a driver of growth, not a detail. Exergy analysis shows that useful work supplied to an economy is Granger-causal to output growth, and the associated research argues that much of twentieth-century economic growth, typically credited to general technological progress in models like Solow's, can be more accurately attributed to tangible improvements in converting primary energy into usable physical work. References: Warr and Ayres (2012), "Useful work and information as drivers of economic growth," Ecological Economics 73, 93–102; Warr, Ayres, Eisenmenger, Krausmann and Schandl (2010), "Energy use and economic development," Ecological Economics 69(10), 1904–1917; Ayres, van den Bergh, Lindenberger and Warr (2013), "The underestimated contribution of energy to economic growth," Structural Change and Economic Dynamics, 79–88. The goods-food-energy triad of Section 2 is therefore not rhetorical; energy sits inside the growth engine itself.
In a crisis, money cannot buy goods that nobody produces. The 2020 pandemic ran the experiment. Eighty countries and customs territories banned or limited the export of face masks, protective gear, gloves and other goods to mitigate shortages, as the World Trade Organization reported. Since the pandemic impacted virtually every country worldwide, there were no unaffected reservoirs of industrial activity to draw from. Dependence was extreme: China accounted for more than eighty-five percent of all United States imports of personal protective equipment and decreased exports, removing critical supplies from markets across the globe; reduced production and exports of pharmaceuticals and personal protective equipment from China exacerbated shortages of critical medical supplies. Money-rich, production-poor countries bid for goods that did not exist.
The 2021 semiconductor shortage repeated the experiment in a different sector. The semiconductor shortage cost the worldwide automotive industry an estimated two hundred and ten billion dollars in lost revenues in 2021, with production of 7.7 million vehicles lost; another estimate holds that more than 9.5 million units of global light-vehicle production were lost in 2021 as a direct result of a lack of necessary semiconductors. Crucially, money was present and goods were not: demand was strong, and together with accelerating prices and sharply declining inventories, the evidence pointed to an insufficient supply of new cars. Governments learned the lesson in the production vocabulary, not the money vocabulary: the United States government offered fifty billion dollars in incentives to Intel, Samsung, and the Taiwan Semiconductor Manufacturing Company to set up factories in the United States. Production is the only insurance policy that pays out in every crisis; money is an insurance policy that pays out only in good weather.
7. THE PROPOSAL: TWO-LAYER NATIONAL ACCOUNTING
Our paper proposes that national statistical offices maintain, alongside the existing monetary accounts, a second official account with two rules.
Rule one, reclassification. Instrumental-layer activity is booked as intermediate cost, never as final output. The instrumental layer is defined operationally: an activity belongs to the instrumental layer when the output of the activity is the coordination of other activities rather than a good or service that a person finally consumes. Finance, insurance, financial trading, tax compliance, and administrative overhead belong to the instrumental layer under the operational definition. The reclassification follows the original logic of Kuznets, Section 3 documented that Kuznets excluded the state from national income because public services were an input for private production whose inclusion would cause double counting, and extends the logic from government services to the entire coordination apparatus of the economy. Under rule one, financial churn can no longer appear as growth, for the same reason that the flour a bakery consumes cannot appear as bread.
Rule two, physical measurement. Productive-layer output is measured directly in physical and near-physical units. Headline indicators include: homes completed and dwelling floor area added; megawatt-hours of electricity generated, with capacity and import dependence stated; tonnes and calories of food produced; potable water delivered; passenger-kilometres and tonne-kilometres of transport provided; semiconductor units fabricated; hospital treatments completed and patients cured; students brought to certified competence. The units are chosen so that each headline answers a citizen-level question: can a family find a home, afford heat, eat well, travel, and be cured?
The method stands on existing statistical practice. Physical accounting is not an invention of our paper; physical accounting is an established branch of official statistics that our paper repurposes from environmental monitoring to the national scoreboard. Economy-wide material flow accounting is an established method to monitor resource use across scales, and the headline indicators of the method are widely used in policy; the framework provides consistent compilations of the material inputs to national economies, changes in material stocks within the economic system, and material outputs, reference: Krausmann, Schandl, Eisenmenger, Giljum and Jackson, "Material Flow Accounting: Measuring Global Material Use for Sustainable Development," Annual Review of Environment and Resources, volume 42 (2017). Deeper still, the physical input-output table records all physical flows associated with the economic activities defined in the System of National Accounts, flows of physical products, extraction, wastes and stock changes, and uses the same classification schemes as the monetary input-output table, so data from the physical table and the monetary table can be combined for analysis and policy modeling. National implementations already exist: comprehensive national input-output tables in purely physical units have been compiled for Germany, Finland and Denmark, beginning with the physical input-output tables of the German Federal Statistical Office (Stahmer, Kuhn and Braun, 1997), and the United Nations handbook on supply and use tables contains a chapter on compiling physical supply and use tables and environmentally extended input-output tables. Statistical offices therefore already possess the table structures; our paper asks statistical offices to promote the physical tables from environmental annex to national scoreboard, and to add the reclassification of rule one.
Measuring intelligence production. The production of intelligence, education, research, engineering skill, resists simple physical units, and our paper treats the weighting of intelligence production as a named open parameter rather than a solved problem. An established accounting literature nonetheless exists: the Jorgenson-Fraumeni approach measures human capital as the present discounted value of the lifetime earnings of all living people in an economy, proposing the increment to lifetime labor incomes as the measure of the output of education and training, and later work integrates Jorgenson-Fraumeni human capital into a complete national income accounting system, revealing a fresh perspective on economic growth across time. Two-layer national accounting can adopt the lifetime-income method as the initial weighting for intelligence production while flagging the money-denominated character of the method as a compromise awaiting better physical proxies (certified competences, completed apprenticeships, peer-reviewed findings).
Differentiation from existing alternatives. Alternatives to gross domestic product exist and are respected: governments and organizations already follow the Human Development Index, the Better Life Index and the Genuine Progress Indicator. The Genuine Progress Indicator, the most economically ambitious alternative, starts with the same personal consumption data as gross domestic product but adjusts for income distribution, adds the value of household and volunteer work, and subtracts costs such as crime and pollution. Every named alternative is therefore a money-denominated welfare adjustment: the alternatives correct the price total. Two-layer national accounting makes a different move: two-layer national accounting refuses the price total as the headline, counts real units on the productive layer, and books the instrumental layer as overhead. The two moves are complementary, a country can publish a Genuine Progress Indicator and a two-layer account side by side, but the moves are not the same move, and the difference is the novelty claim of our paper.
What our paper does not change. Two-layer national accounting changes the scoreboard of the state; two-layer national accounting does not abolish market prices, does not replace private bookkeeping, and does not plan production. Firms keep trading at market prices; the monetary national accounts continue to be compiled; the physical account and the monetary account share classifications by design, as the physical input-output literature already demonstrates. Section 10 returns to the point when answering the central-planning objection.
8. ANTICIPATED FINDINGS, EVALUATION PROCEDURE, AND OPEN PARAMETERS
Our paper is a position paper with a defined empirical program. Three findings are anticipated.
Anticipated finding one, rankings shift. When instrumental-layer activity is reclassified as intermediate cost, countries with large financial centers fall relative to the money-denominated rankings of the same countries, and countries with strong physical output per person rise. The size of the shift measures how much of the recorded prosperity of each country is coordination overhead rather than production.
Anticipated finding two, ministry incentives flip. Where ministries are evaluated on physical results, homes completed, energy generated, patients cured, students competent, spending changes sign in the evaluation: spending becomes a cost to minimize rather than an achievement to report. Nobody praises a bakery for the flour the bakery consumed; under two-layer national accounting, no ministry is praised for the budget the ministry consumed.
Anticipated finding three, production predicts crisis survival. Countries ranking high on the productive-layer scoreboard before 2020 experienced smaller shortage shocks during the pandemic and the semiconductor crisis than countries ranking high on money-denominated scoreboards, because physical capacity, the thing the new scoreboard counts, is the thing that shortages test.
Evaluation procedure. The procedure has four steps and uses only published statistics, so any statistical office or research team can replicate the procedure. Step one: for a pilot set of five to ten countries, rebuild national accounts under rule one, using existing supply-and-use tables to identify instrumental-layer value added and reclassify the identified value added as intermediate consumption. Step two: assemble the physical headline indicators of rule two from existing material flow accounts, energy balances, construction statistics, agricultural statistics, and health statistics. Step three: compare country rankings and trend lines under the monetary account and under the two-layer account, and measure the divergence. Step four: run the two historical event studies, the 2020 medical supply shock and the 2021 semiconductor shock, and test whether the two-layer scoreboard, computed with pre-crisis data only, predicts crisis performance better than gross domestic product predicts crisis performance.
Named open parameters. First, the weighting of intelligence production and culture production (Section 7 states the candidate method and the reservation about the candidate method). Second, quality adjustment: a home completed in 2026 is not a home completed in 1970, and hedonic correction of physical units imports prices through the back door; the degree of permitted hedonic correction is a design dial, not a settled constant. Third, boundary assignments: some activities serve both layers (software, logistics, wholesale trade), and the classification manual must assign the dual activities by final-use tests, with sensitivity analysis published for contested assignments. Fourth, aggregation: whether the physical headlines remain a dashboard or are compressed into a small number of indices, and by which weights, our default position is the dashboard, consistent with the recommendation of the Stiglitz-Sen-Fitoussi Commission to move towards a broader dashboard of indicators.
9. IMPLICATIONS FOR GOVERNMENT POLICY
9.1 The metric flip inside ministries. The most direct application of two-layer national accounting is administrative: evaluate public organizations on productive-layer results. The evidence on public-sector targets is two-sided, and our paper uses both sides. The cautionary side: Bevan and Hood (2006), "What's Measured Is What Matters: Targets and Gaming in the English Public Health Care System," Public Administration, 84, 517–538 documents how target regimes invite gaming, summarized as improvements in reported performance accompanied by attempts to game the system, such as "hitting the target and missing the point". The constructive side: targets with sanctions can work when the targets accurately measure key dimensions of performance that cannot easily be manipulated by gaming, examples include hospital waiting times and ambulance response times, and in the English natural experiment by 2002 hospital waiting times had decreased in England but increased in Wales. The design lesson for two-layer national accounting is exact: choose final physical outcomes (a completed home, a cured patient) rather than clerical proxies, publish several indicators together rather than one, and audit on randomized schedules. Physical finality is the gaming resistance.
9.2 Aligning the treasury with production. A companion paper in our research program proposes replacing complex taxation with a flat levy collected automatically on each transaction through a central-bank payment system, at a rate consistent with historically persistent flat levies. Under the combined design, government revenue grows only when real transactions grow, so the treasury acquires a direct structural interest in expanding production and exchange rather than inflating asset prices. Two-layer national accounting supplies the scoreboard; the transaction levy supplies the incentive; the two instruments point the state at the same target, the productive layer.
9.3 Supply expansion is the silent raise. The cheapest way to make citizens richer is to make the goods citizens buy more plentiful and therefore cheaper, and housing is the leading case. Glaeser and Gyourko (2018), "The Economic Implications of Housing Supply," Journal of Economic Perspectives 32(1), 3–30 show that the gap between house price and production cost can be understood as a regulatory tax, and the implicit tax on development created by housing regulations is higher in many areas than any reasonable negative externalities associated with new construction. A government watching money flows sees rising house prices as wealth; a government watching the productive layer sees a regulatory tax on the young and a construction deficit to close. The same logic covers energy: megawatt-hours generated, not energy-sector revenue, is the number that predicts whether factories run and homes stay warm, and Section 6 established that useful energy is a driver of growth itself.
9.4 Health: buying cures instead of renting health. Money-flow metrics reward permanent treatment; production metrics reward cure. The finance literature states the distinction plainly: taking traditional medications is like renting an apartment, paying for the benefit a month at a time, while curative therapies are akin to buying a house, and the difference is between "renting health" one pill at a time and buying a cure; reference: Montazerhodjat, Weinstock and Lo (2016), on financing curative therapies, Science Translational Medicine. The market's own analysts confirmed the incentive problem: a 2018 Goldman Sachs biotechnology report asked, "Is curing patients a sustainable business model?", observing that hepatitis C treatments achieving cure rates above ninety percent saw United States sales peak at 12.5 billion dollars in 2015 and fall steadily thereafter because curing patients means no recurring revenue and, for infectious disease, shrinks the pool of new patients. Under two-layer national accounting the paradox dissolves: the health system is scored on patients cured, a productive-layer output, so public subsidy and public procurement flow toward cures, prevention, and rejuvenation, and away from permanent symptom subscriptions. Citizens get healthier; recorded money flows in the health sector may well shrink; under the new scoreboard, shrinkage of overhead while cures rise is registered as success, which is the whole point.
9.5 Attacking the ratchet of the state share. Two robust regularities explain why the instrumental share of national life ratchets upward. First, Wagner's law, the law of increasing state activity, observes that public expenditure increases as national income rises, with modern panel support in bootstrap Granger-causality results favoring Wagner's law in seven of twelve OECD countries, carrying the policy implication that aging populations and welfare demand may force higher taxes or excessive borrowing. Second, the Baumol effect: wages in jobs with little productivity growth rise in response to rising wages in progressive sectors, so services, health, education, care, administration, become relatively more expensive over time, a mechanism with strong professional consensus: when weighted by confidence, eighty-eight percent of a polled panel of United States economic experts agreed that rising manufacturing productivity leads the cost of labor-intensive services to rise. The overhead is not small: compliance with the United States federal tax code consumes 7.9 billion hours plus out-of-pocket costs, totaling roughly 546 billion dollars per year, nearly two percent of gross domestic product; Canadian estimates place private business compliance costs between 1.2 and 1.8 percent of gross domestic product, with public administration adding 0.2 to 0.5 percent; and the burden is regressive: as a proportion of taxes paid, compliance costs of medium-sized businesses are significantly higher than for the largest businesses (see also Slemrod, "Tax Compliance and Enforcement," Journal of Economic Literature 57(4), 904–954, 2019). Two-layer national accounting makes the ratchet visible year by year, the instrumental layer is a published overhead ratio, and administrative automation becomes the one form of automation with no demand-side downside, because bureaucracy has no consumers to lose: every bureaucrat-hour freed is pure gain for the productive layer.
9.6 Strategic capacity before the crisis, not after. Section 6 recorded that governments responded to the semiconductor shock with fifty billion dollars of factory incentives, production policy adopted in panic, after the shortage. A productive-layer scoreboard publishes chip fabrication, food self-sufficiency, energy capacity, and medical manufacturing continuously, so legislators see the vulnerability while repair is still cheap. Under money-flow accounting, resilience is invisible until the day resilience is priceless.
10. OBJECTIONS AND ANSWERS
Objection one: "Physical targets are Soviet planning." The objection deserves a direct answer. Soviet material planning replaced market prices, private property, and consumer choice with administrative production orders; the failure of Soviet material planning is documented, and the target-gaming pathologies of Soviet planning are the canonical illustrations of Goodhart's law, indeed the public-administration literature draws the parallel explicitly, noting that the English targets regime had obvious parallels with the Soviet regime, resting on synecdoche, taking a part to stand for the whole, and on the assumption that measurement problems and gaming do not matter. Two-layer national accounting shares no instrument with Soviet planning: markets keep setting prices, firms keep choosing outputs, consumers keep choosing purchases, and the monetary accounts keep being published. The only thing that changes is the scoreboard by which citizens judge governments and governments judge ministries. A nation that counts homes is not a nation that orders homes; a nation that counts homes is a nation that notices when homes are missing.
Objection two: "Goodhart's law will corrupt your physical metrics too." Correct as a warning, and Section 9.1 already contains the answer: the empirical record shows that gaming concentrates where measures are clerical proxies, while targets measuring key dimensions that cannot easily be manipulated, waiting times, response times, produced real improvement. Two-layer national accounting therefore selects final physical outcomes, publishes dashboards rather than single numbers, randomizes audit timing, and treats every metric as reviewable. Goodhart's law counsels careful metric design; Goodhart's law does not counsel steering by a corrupted metric because all metrics are imperfect.
Objection three: "Prices aggregate dispersed information; physical units discard the information." The Hayekian point about prices is accepted in full, for market coordination. Firms and households should keep using prices, and under our proposal firms and households do keep using prices. The claim of our paper is narrower: the aggregate of money flows is a corrupted summary statistic for national success, because Section 5 showed that money flows decouple from production at the aggregate level. Nothing in the price system requires that the state's headline scoreboard be a money total; the physical input-output literature demonstrates that physical tables and monetary tables share classification schemes and are combined for analysis, the two informations coexist by construction.
Objection four: "You cannot add homes to megawatt-hours to cured patients." Correct, and our paper does not add homes to megawatt-hours to cured patients. The productive-layer account is a dashboard of headline physical indicators, following the dashboard recommendation of the Stiglitz-Sen-Fitoussi Commission. Aggregation weights, where a compressed index is politically demanded, are a named open parameter with published sensitivity analysis. A dashboard that refuses false precision is more honest than a single number that hides the refusal.
Objection five: "Finance matters; your proposal starves finance." Our paper claims calibration, not abolition. The threshold evidence cuts both ways and our paper cites both ways: the level of financial development is good up to a point, finance below the threshold serves production, and credit that funds a factory passes the service test of Section 2. Two-layer national accounting simply prices the instrumental layer as what the instrumental layer is, a cost, so a nation can ask continuously whether each additional unit of coordination overhead buys additional production. Tobin asked the question once in 1984; the scoreboard asks the question every year.
Objection six: "Existing indices already fix gross domestic product." Section 7 answered the objection in detail: the existing indices are money-denominated welfare adjustments, while two-layer national accounting is a physical-unit scoreboard plus an intermediate-cost reclassification, reviving the original classification practice of Kuznets. Complementary, not identical.
Objection seven: "Physical units ignore quality and new goods." Partly correct, and stated as an open parameter: hedonic and quality adjustment of physical units is permitted by design dial, and the intelligence-production account leans on the lifetime-income measure of education output until better physical proxies mature. Our paper accepts a rough measure of the right thing over a precise measure of the wrong thing.
11. LIMITATIONS
Our paper carries five limitations, stated plainly. First, culture production and care work resist both money measures and physical measures; the Genuine Progress Indicator tradition of adding the value of household and volunteer work marks the gap without closing the gap, and two-layer national accounting inherits the gap. Second, the reclassification boundary of the instrumental layer requires a classification manual with contested cases, and classification manuals attract lobbying; the defense is publication of sensitivity analyses, not the pretense of a lobby-proof manual. Third, international comparability requires standardization work on the scale of the existing System of National Accounts revisions, a decade-class project, honestly stated. Fourth, the evaluation procedure of Section 8 tests predictive validity on two crises; two crises are suggestive, not conclusive. Fifth, the deepest obstacle is political rather than technical: policy makers and politicians of all political preferences continue to give much importance to gross domestic product, and a scoreboard that reclassifies large industries as overhead will be resisted by the industries so reclassified. Our paper's answer to the fifth limitation is the whole argument: the measurement corruption is precisely why the correction must live in official statistics rather than in discretionary politics.
12. CONCLUSION
Money is not the economy. The economy is farms, factories, power plants, homes, hospitals, and schools, the productive layer, and money is the instrumental layer that helps the productive layer coordinate. Gross domestic product, a money total, was introduced with a warning from the inventor of gross domestic product, "The welfare of a nation can scarcely be inferred from a measurement of national income", and ninety years of measurement corruption, financialization, and crisis evidence have proven the warning. Our paper proposes the correction the inventor himself began: reclassify coordination as cost, count production in real units, and let governments be judged by homes completed, energy generated, food produced, and patients cured. The instruments exist in the statistical offices already; the evidence base exists in the journals already; the missing element is the decision. A country that adopts the two-layer scoreboard does not become a planned economy; the country becomes a seeing economy, and a country that sees production clearly will choose production, because production, not money, is what keeps citizens alive, housed, warm, and well. The purpose of a country is the citizens of the country.