Steering the Economy with a 1940 Map
GDP, in its original conception, was developed in the 1930s to address the Great Depression. Its standardized form, used to plan the war effort, emerged in the early 1940s. Simon Kuznets, the economist who laid its foundations, himself warned against using this indicator as a measure of national well-being. Nobody listened. Eighty years later, this tool steers policies on health, education, climate transition, and digital regulation in economies that Kuznets would not have recognized.
The Measure of Progress, published this year by Diane Coyle, is not a screed against growth. It is a precise diagnosis of what our measurement instruments capture, what they ignore, and why this structural blindness costs increasingly more as the real economy moves away from the industrial model on which national accounting is based.
The Essential Points
- GDP, constructed in the 1930s-1940s to measure industrial production, structurally excludes several major sectors of the contemporary economy: zero-price digital services, unpaid care work, environmental degradation, and data externalities.
- Diane Coyle, professor at Cambridge and former adviser to the British Treasury, documents the extent of the bias: real productivity growth would be systematically underestimated in economies with a strong digital component.
- Attempts at statistical reform encounter well-documented institutional resistance: changing the indicator amounts to changing political priorities, and the beneficiaries of the current framework are not neutral in this debate.
- Over thirty years, if the gap between measured wealth and real wealth continues to widen, public policies risk being calibrated on an image of the economy increasingly unfaithful to the lived experience of households.
The Author
Diane Coyle teaches economics at the University of Cambridge, where she directs the Bennett Public Policy Chair. She has served on the Migration Advisory Committee and was Senior Economic Assistant at HM Treasury from 1985 to 1986, as well as informal adviser to the Treasury at other points in her career. A consultant, columnist, and founder of the Enlightenment Economics consulting firm, she works at the intersection of growth economics, public policy, and statistical measurement. Her previous book, GDP: A Brief but Affectionate History (2014), had already laid the groundwork for this reflection. The Measure of Progress deepens ten years later the diagnosis, integrating the transformations produced by digital technology, climate crisis, and pandemic.
What GDP Does Not See
Coyle’s central thesis can be stated in one sentence: our macroeconomic indicators measure an economy that no longer quite exists. Not because they would be poorly constructed by the standards of their time, but because the economy has changed in nature and the instruments have not kept up.
She identifies three principal blind spots.
The first is zero-price digital services. When a household devotes three hours a day to services provided free in exchange for personal data, GDP records zero value created. Yet these services exist, they are used, they sometimes replace paid market services. Erik Brynjolfsson and his MIT colleagues attempted to quantify this invisibility in a series of works published between 2019 and 2023: according to their estimates, American consumers would value access to the internet and associated free services at a considerable sum per year. This figure enters into no national accounts.
The second blind spot is unpaid work. Care for children, the elderly, the sick, domestic work: according to the British Office for National Statistics, this work represented in 2016, valued at the cost of market replacement, the equivalent of 63.1% of the UK’s GDP. Since the pandemic, demand for informal care has increased in most OECD countries, notably due to demographic aging. This appears nowhere in indicators of national economic performance.
The third blind spot is environmental. National accounting records the cutting of a forest as an addition to GDP. Degradation of the underlying natural capital is not subtracted. The Stiglitz-Sen-Fitoussi Commission, mandated in February 2008 by Nicolas Sarkozy with its final report submitted in September 2009, had already documented this asymmetrical treatment. Fifteen years later, it remains intact in the vast majority of official statistics, despite work on natural capital accounting conducted by the World Bank and the Partha Dasgupta Review published in 2021 for the British government.
Why Reform Does Not Advance
Coyle does not simply map out the gaps. She tackles a more difficult question: if the academic consensus on GDP’s limitations is both broad and old, why do national statistical systems move so little?
Her answer is institutional, and it is convincing. Indicators are not neutral. They structure political priorities, budgetary trade-offs, eligibility criteria for public assistance. A government that integrates environmental externalities into its growth measurement mechanically ends up with weaker growth and a more visible climate action constraint. A government that recognizes the value of care work must then answer the question of its compensation. Changing the indicator amounts to changing what the state sets as its objective — and thus to redistributing power among actors benefiting from the existing framework.
This is not a conspiratorial thesis. It is ordinary political science observation. National statistical institutes depend budgetarily on the governments they serve. Major methodological reforms require long and costly international coordination. And alternative indicators proposed over thirty years — the Human Development Index, Gross National Happiness index, subjective well-being indicators, “Beyond GDP” — struggle to achieve the operational clarity of GDP, which remains, despite everything, simple, comparable over time and between countries.
Coyle does not argue for abandoning GDP. She argues for a dashboard: complementary indicators, constructed with the same methodological rigor, for dimensions that GDP deliberately ignores. This position is more modest than it seems — and perhaps that is what makes it more robust.
The Productivity Paradox, Digital Version
A central chapter of the book returns to what Coyle calls the measurement paradox in the digital age. American productivity, measured by standard methods, stagnated between 2005 and 2015, while the diffusion of smartphones, the cloud, platforms, and collaboration tools accelerated at unprecedented speed in recent technological history. Robert Solow had already observed something similar with computing in the 1980s: “You can see the computer age everywhere but in the productivity statistics.”
Coyle’s hypothesis is that the paradox says as much about our instruments as about our economy. If a growing fraction of value created circulates outside measurable monetary transactions, statistical productivity can stagnate while real productivity advances. This hypothesis is difficult to settle empirically, but it is taken seriously by a growing number of economists. The work of Chad Syverson (University of Chicago), notably his 2017 article in the Journal of Economic Perspectives, showed that the magnitude of digital measurement bias probably did not suffice alone to explain the observed productivity slowdown, but that its contribution remained significant and underestimated.
The pandemic added an additional layer of ambiguity. Massive remote work produced gains in unmeasured commuting time, a reorganization of consumption spending, an acceleration of digital adoption. All of this translated into volatile GDP figures difficult to interpret, and into statistical debates among institutes on the method of valuing public services whose production had shifted online. Coyle documents these debates with a precision that owes much to her institutional experience.
The Tension with Market Liberalism
One must name here a tension that the book explores without always resolving frankly. Tyler Cowen, an economist of secular stagnation whose work regularly crosses paths with Coyle’s, has argued a neighboring thesis but with inverted conclusion: if the digital economy creates invisible value, this means we live better than our indicators show, and the experienced stagnation is partly a statistical illusion. This is an optimistic reading of the same diagnosis.
Coyle does not dismiss it, but responds with a governance argument. Even if real growth exceeds measured growth, public policies are calibrated on measured growth. Health budgets, social transfers, infrastructure investment decisions follow rules based on obsolete aggregates. The statistical gap is not merely academic: it produces policies undersized for real needs. This shift from economic argument to political argument is what distinguishes The Measure of Progress from specialist debate.
There is another tension, more rarely raised. Data, in the digital economy, is both private asset and potential public good. Its statistical valuation poses questions of ownership and access that exceed national accounting. The open data revolution in science has shown what opening resources could produce in collective gains; the same logic applies to economic data. But an indicator that values the data assets of platforms risks reinforcing their dominant position rather than shedding light on their net social contribution. Coyle raises the problem without entirely settling it, which is honest.
The Long Arc: Thirty Years of Possible Drift
The prospective dimension of the book is perhaps the most unsettling. Coyle does not formulate it as prophecy; she formulates it as a conditional trajectory. If statistical systems do not reform within the next two decades, the gap between measured wealth and real wealth will become structural.
This trajectory has several drivers. Demographic aging will increase the weight of unpaid care work in developed economies. The energy transition will create value exchanges outside conventional market circuits — energy savings, ecosystem services, climate resilience. Artificial intelligence will displace the boundary between paid work and automated work in ways that current categories of national accounting cannot grasp. AI agents that already produce scientific results in hours are reshaping the contours of measurable intellectual production.
The risk is not abstract. Health policy calibrated on GDP that ignores care work will structurally underfund the sector. Industrial policy that does not value the positive externalities of digital technology will underinvest in the infrastructures that make them possible. Climate policy based on GDP that ignores natural capital degradation will systematically defer difficult trade-offs. This is not a hypothesis: it is the mechanism that Coyle documents by following the history of British budgetary decisions since the 1980s.
There is also a generational dimension that the book addresses too briefly. Current generations inherit degraded natural capital, public debt partly incurred to finance immediate spending at the expense of long-term investment, and a digital economy whose negative externalities — concentration of market power, erosion of privacy, destabilization of labor markets — have never been properly accounted for. Measuring these intergenerational transfers requires precisely the instruments that GDP does not provide.
The Book’s Own Blind Spots
The Measure of Progress has its own limitations, which Coyle might benefit from stating more directly.
The first is an unresolved political tension. The proposal for a dashboard of complementary indicators assumes political consensus on dimensions to measure. Yet this consensus is precisely what is lacking. What weight to give to equality, sustainability, subjective well-being relative to market production? These weightings are political choices, not technical choices. Coyle tends to treat this question as a statistical design problem when it is also a problem of political philosophy.
The second limitation is geographical. The book is centered on developed, primarily English-speaking economies. The measurement challenges for middle and low-income countries are different: the informal economy represents according to the ILO more than 60% of total global employment, an even higher proportion in low-income countries. A statistical tool designed for American or British digital services does not solve the measurement problem in these contexts. This limitation does not disqualify the book, but it restricts the scope of its recommendations.
The third limitation is practical. Coyle argues for methodological rigor in new indicators, but the examples she cites — subjective well-being surveys, environmental satellite accounts — remain today marginal in real political decisions. The explanation she gives for this resistance (institutional, political) is convincing, but it does not say much about how to escape it. The diagnosis is solid. The theory of change is thinner.
Why Read This Book
The Measure of Progress is addressed to anyone who takes public policy seriously — economists, institutional decision-makers, journalists, engaged citizens. It does not demand particular technical training. It does require suspending the obviousness of GDP as a natural measure of economic progress, which is intellectually more difficult than it appears.
What one finds here that one does not easily read elsewhere is the combination of rigorous economic analysis and a refined institutional reading of resistances to change. Coyle does not make statistical measurement a technical problem to be solved by experts. She makes it a problem of democratic governance: whoever decides what gets measured decides, in practice, what counts as important.
At a moment when debates on growth, state financing, climate transition, and the future of work occupy the center of political agendas — as economic tensions that global fragmentation is awakening testify — this reminder has direct relevance. Measurement instruments are never neutral. They manufacture an image of the economy, and this image orients policies. Knowing what it leaves in the shadows is not academic curiosity. It is a basic condition for steering correctly.
Bibliographic Information
Title: The Measure of Progress
Author: Diane Coyle
Publisher: MIT Press
Year of Publication: 2025
Pages: 272 p.
Sources
- Diane Coyle, The Measure of Progress, MIT Press, 2025 — Related to Institut Montaigne publication
- Brynjolfsson, Eggers, Grindal, “Using Massive Online Choice Experiments to Measure Changes in Well-being”, PNAS, 2019 — MIT Initiative on the Digital Economy
- Office for National Statistics (UK), “Household Satellite Account”, 2016 — ons.gov.uk
- Partha Dasgupta, The Economics of Biodiversity: The Dasgupta Review, HM Treasury, 2021 — gov.uk
- Stiglitz-Sen-Fitoussi Commission, Report on the Measurement of Economic Performance and Social Progress, 2009 — vie-publique.fr
- Chad Syverson, “Challenges to Mismeasurement Explanations for the U.S. Productivity Slowdown”, Journal of Economic Perspectives, 2017
- International Labour Organization (ILO), Women and Men in the Informal Economy, 2023 — ilo.org
- ONS Household Satellite Account 2016 (figure 63.1% of GDP) — ons.gov.uk
- MIT Press – The Measure of Progress (2025) — mitpress.mit.edu
- AEA – Syverson (2017) on mismeasurement — aeaweb.org
- Official CV Diane Coyle – Bennett School — bennettschool.cam.ac.uk
- Wikipedia – Commission on the Measurement of Economic Performance and Social Progress — en.wikipedia.org
- MIT Sloan – Brynjolfsson Study on Digital Services (2019) — mitsloan.mit.edu
- Wikiquote – Kuznets (Report to Congress, 1934) — en.wikiquote.org