Numéraire · The MeasuresThe Contest Inside the Number
Numéraire
A Companion to the Atlas
The Measures
the contest inside the number
the book has said that every price is a verdict on a hidden contest; here the same is shown of every official statistic — the figures economic life actually runs on
Why the numbers are here
Every other companion argues about concepts. But economics runs on numbers — GDP, inflation, unemployment, inequality — presented as the most neutral facts there are. Each is in truth built, and the building hides a choice, and behind the choice is a contest. The figure is hidden here at first, as the news presents it. Reveal the contest inside.
fifteen measures · what each claims, how each is built, and the contest each conceals
On the numbers we trust
Every Statistic a Verdict
The book has made one argument, over and over, in twenty-six forms: that an economic quantity is never a simple fact of nature, but a construction that conceals a contest. It has made that argument about concepts — the wage, the rate, the share. It has not yet made it about the place the argument matters most: the official numbers. GDP, the inflation rate, the unemployment figure, the measure of inequality — these are the statistics that move markets, set pensions, decide elections, and define who counts as poor. They are reported in the flat, authoritative voice of arithmetic, as though they were simply read off the world. They are not. They are built, and every choice in the building has a winner and a loser.
Consider what it takes to make one. To measure inflation you must first choose a basket of goods, and a basket is a guess about whose life is typical. To count the unemployed you must first decide who counts as looking for work. To put a price on a ton of carbon you must first decide how much a harm to the unborn is worth today. None of these is a measurement; each is a judgment that the measurement then hides. The number that emerges carries the judgment silently into every decision that consults it — which is the quiet genius, and the quiet danger, of a statistic: it settles an argument without anyone noticing the argument was had.
Below are fifteen of the measures economic life runs on, each with what it claims to measure, how it is in fact built, and the contest hidden inside it. The contest is folded away at first, so that each number appears as it does in the news — clean, settled, beyond dispute. Open them one at a time, or reveal them all at once. The point of the exercise is not to discredit the figures, which are indispensable, but to learn to read them: to see, in each one, both the quantity it reports and the verdict it encodes.
Gross Domestic Product
the size of the economy
Claims to measurethe total market value of all the final goods and services a nation produces in a year.
How it is builtby adding up consumption, investment, government spending, and net exports — everything bought and sold, valued at its market price.
The contest inside the number
It counts only what carries a price. The unpaid work that holds a society together — the childcare, the eldercare, the cooking, the labor done overwhelmingly by women at home — is worth nothing to GDP, and becomes valuable only the instant it is sold. It treats destruction as production: an oil spill raises GDP through the cleanup, and the felling of a forest counts as income while the loss of the forest counts as nothing. It cannot tell a dollar of weapons from a dollar of vaccines. What a nation agrees to call “the economy” is the decision of what labor will be seen and what will be made invisible.
Claims to measurehow much the prices faced by a typical household rise from one year to the next.
How it is builtby pricing a fixed basket of goods, weighted by average spending, and adjusting for changes in quality and for shoppers switching to cheaper substitutes.
The contest inside the number
There is no typical household. The poor spend far more of what they have on rent, food, and energy — the very things that often rise fastest — so a single index understates inflation for those least able to bear it. And the technical adjustments, for quality and for substitution, almost all push the measured number downward; since pensions, the minimum wage, tax brackets, and benefits are all tied to that number, a lower reading means quietly smaller raises for millions. The choice of basket is a choice about whose cost of living counts.
Claims to measurethe share of people who want a job and cannot find one.
How it is builtfrom a household survey that counts you as unemployed only if you are out of work and have actively looked within the past four weeks.
The contest inside the number
The four-week rule decides who exists. Those who have given up looking — the discouraged — vanish from the count, along with those working part-time who want full-time, and those who have dropped out of the labor force altogether. The headline rate can fall not because anyone found work but because more people stopped searching for it. Where the boundary of “the labor force” is drawn decides who is even permitted to count as wanting a job.
Claims to measurethe earnings of a typical worker.
How it is builtusually as the mean — total wages divided by the number of workers — or sometimes the median, the pay of the worker exactly in the middle.
The contest inside the number
The mean and the median tell opposite stories. The mean is dragged upward by a handful of enormous incomes, so “average pay” can rise for years while the worker in the middle gains nothing at all. Worse, the average can climb for perverse reasons: when low-wage workers lose their jobs in a downturn, the average of those who remain goes up, and the figures report a raise that no one received. The choice between mean and median is the choice between flattering the economy and describing most of the lives inside it.
Claims to measureinequality in a single number, from zero, where everyone is equal, to one, where one person has everything.
How it is builtfrom the gap between the actual distribution of income and a line of perfect equality.
The contest inside the number
One number cannot hold a distribution. A society where the top one percent seizes everything and one where the upper-middle merely pulls away can share the same Gini; the measure is least sensitive precisely at the extremes — the very rich, the very poor — that matter most. And the answer changes entirely with the fine print: income before tax or after, individuals or households, wages alone or wages plus benefits. A single figure of inequality conceals which inequality is meant.
Claims to measurethe income below which a household cannot afford the necessities of life.
How it is builtin the United States, from a calculation of the 1960s — three times the cost of a minimum food budget — updated only for inflation in every year since.
The contest inside the number
It is a political decision wearing the clothes of a measurement. Poverty is relative and many-sided, but the line is a single absolute number frozen in an old assumption; draw it a dollar higher or lower and millions move into or out of “poverty,” and with them the right to assistance. The line does not discover the poor. It defines them — and it decides how many a society is willing to see.
Claims to measurehow efficiently an economy turns labor into goods.
How it is builtby dividing output, a slice of GDP, by the hours of work it took to produce.
The contest inside the number
It inherits every blind spot of GDP, and adds one of its own: it credits labor for gains that often belong to the machine, the technology, or simply to people being worked harder. “Rising labor productivity” can mean the worker is more exhausted, not better off — and for decades productivity has climbed steeply while the typical wage has flattened, severing the very link to living standards the measure is forever invoked to prove.
How it is builtas a single headline figure — the rate a central bank sets, or a benchmark a government pays.
The contest inside the number
There is no such thing as the interest rate. The rate a central bank announces, the rate a blue-chip firm pays, and the rate a payday borrower is charged differ by orders of magnitude; to speak of “the” rate is to hide that the price of credit is steeply tiered by power, collateral, and risk — cheap for those who least need it, ruinous for those who most do. And a rate means nothing until it is set against inflation: five percent while prices climb six is a loss dressed up as a gain.
Claims to measurehow much turbulence the market expects — the figure often called the fear gauge.
How it is builtfrom the prices investors pay for options that protect against swings in a stock index.
The contest inside the number
It measures not danger but the expectation of danger, and only as priced by those wealthy enough to be in the options market. It is lowest exactly when complacency is greatest, and it spikes only once the fear has already arrived — so it tends to confirm a panic rather than to warn of one. A number sold as a measure of risk is really a measure of a particular crowd’s nerves.
Claims to measurethe total quantity of money in the economy.
How it is builtas a series of tallies — M1, M2, and others — that add up cash, then checking accounts, then savings and money-market funds.
The contest inside the number
Money has no natural edge. Cash shades into deposits, deposits into savings, savings into instruments that are almost-but-not-quite money, and the measures simply draw lines across that continuum — lines so arbitrary that central banks add them, redefine them, and abandon them. What counts as money is a definitional choice, and the choice alone can make the money supply appear to explode or to hold perfectly still.
Claims to measurethe share of income a person or a company actually pays in tax.
How it is builtas tax paid divided by income — to be sharply distinguished from the statutory rate written into the law.
The contest inside the number
The gap between the statutory rate and the effective rate is where power lives. Headline rates are set in public and debated loudly; the deductions, credits, shelters, and definitions that lower the rate actually paid are written quietly, and they bend most for those who can hire the lawyers to find them. Which income even counts is itself contested — realized or accrued, wages or capital gains — so the wealthiest can face a lower effective rate than the salaried worker. The number you are shown is rarely the number that is paid.
Claims to measureroutinely treated, in the news, as a barometer of the whole economy.
How it is builtby tracking the combined share prices of a set of large corporations.
The contest inside the number
It measures the wealth of shareholders, who are overwhelmingly the richest households — the top tenth owns the vast majority of all stock — so it is a gauge of capital’s fortunes, not of wages, of jobs, or of the economy at large. The index can soar in the very year that real pay falls and joblessness rises. To report the market as though it were the economy is to define the economy, quietly, as the interests of those who own it.
Claims to measurehow competitive a market is, often summarized in a single index of concentration.
How it is builtby squaring and summing the market shares of the firms in an industry — the higher the figure, the fewer the hands.
The contest inside the number
The measure depends entirely on how the market is drawn. Define it narrowly and a firm looks dominant; define it broadly and the same firm looks small — and the drawing of that boundary, fought over in every antitrust case, decides whether power is seen or denied. A single concentration figure also says nothing of the power that runs along a supply chain, or the grip a platform holds over the sellers who depend on it. What we call competition often depends on where we agree to stop looking.
Claims to measurethe dollar damage done by emitting one more ton of carbon dioxide.
How it is builtby modeling the future harm from warming — to crops, coasts, health, and productivity — and translating it into money today.
The contest inside the number
Almost every number that matters in it is a choice. The discount rate — how much a harm to the unborn is worth to us now — can swing the figure many times over on a single assumption, turning a catastrophe into a rounding error and back again. Whose suffering counts, and at what rate a life in one country is traded against a life in another, are decisions made inside the model and hidden by its output. A measure built to price an externality cannot escape being a verdict on how much the future, and the distant, are worth.
Claims to measurehow optimistic households feel about the economy, and so how they are likely to spend.
How it is builtfrom surveys that ask people to judge their finances and their prospects, distilled into an index.
The contest inside the number
It measures a feeling, and feelings are made — shaped by the headlines, by the political weather, and by the very figures above as they are reported. The mood it captures can diverge sharply from any material fact: people may feel poor in a boom or secure on the eve of a bust. Treated as a leading indicator, it often does no more than reflect the story the other numbers have already told — a measure of sentiment mistaken for a measure of the world.
Run through these and a single lesson repeats: not one of these numbers is found. Each is built, and the building embeds a choice — which basket, which boundary, which discount rate, the mean or the median, before tax or after — and behind each choice is a contest with winners and losers. The figures that move markets, set benefits, and decide elections are verdicts, every one, handed down in the voice of arithmetic.
But the sharper lesson is in what has no measure at all. There is no official statistic for the value of unpaid care, for the worth of a stable climate, for the depth of a person’s wants, for the health of a commons, for the trust that lets a market work, or for power itself except where an antitrust case drags a number into being. The quantities this book has cared about most are, again and again, precisely the ones no ledger attempts. That is not an oversight. We manage what we measure and we measure what we can price, so the unpriced and the unmeasured fall out of every decision that consults the figures — which is to say, out of nearly every decision that gets made. The absence of a measure is itself a verdict, and the quietest one of all.
None of this is a case against measurement. A modern economy cannot be seen, let alone governed, without its statistics; the answer to a bad number is a better number and an honest account of its limits, not a retreat into impressions — that way lies the trap of consumer confidence, sentiment mistaken for knowledge. The measures are indispensable, and they are constructions, and the two facts have to be held at once. To read a statistic well is to see both the quantity it reports and the contest it encodes — to remember that the figure in the headline is the end of an argument, not a fact of nature.
It is the book’s oldest claim, arrived at last at the level of the data. Every price is a verdict on a hidden contest; so is every statistic built from prices. The number is never the end of the inquiry. It is the place the inquiry has been told to stop. The argument — over what to count, and over all that the counting leaves out — remains open.
Return to the atlas for the table these numbers measure, the preface for the argument they carry into the world, the constellation for the bonds among them, the properties to weigh them by trait, the compounds for what they form together, the diagonals to read one idea across the table, the objections to hear the case against the book, the chronology for the order in which it was all found, the laboratory to set the models moving, the guided tour for a path through the whole, the index to find any term by name, the cases to see the table at work on real questions, the priceless for what should not be priced at all, the gaps for what is not on the table, the first person for the life behind the numbers, the schools for how each tradition reads them, the sources for the studies behind the numbers, or the unsettled for the contests of our moment. The quantities are everywhere; the contests inside them are the book.