the still engravings made operable — move the inputs, and watch the quantity re-form before you
Why these move Throughout the volumes the figures stood still, each frozen at a single setting. Here a few are made live, so that the book’s one claim may be not asserted but operated: that every number is an outcome, and would be otherwise if the contest were otherwise.
A picture can assert that a price is contingent; only an instrument can prove it. The volumes argued, plate after plate, that the quantities of economic life are verdicts — settlements of contests that might have gone another way. What follows turns thirteen of those arguments into machines. Each is a model the volumes drew as a static engraving, here given controls: take hold of the inputs, move them, and watch the number the model produces re-form in the same instant.
They are arranged to follow the atlas in miniature — from the formation of a price, through the logic of the margin and the abuses of power, to a market’s failure, the pricing of time, the force of growth, the crookedness of the mind, and the trap of strategy — and, in a second movement, the choosing mind, the price of risk, the failure that information breeds, the wedge of a tax, and the shape of a society’s income. Each demonstration names the plate it belongs to, and its title links there, so that the reading and the operating may be done together: read the plate to learn what the model means, then return here to make it move.
None of this is decoration. To drag a supply curve and see the price chase it, to push production past the point where benefit meets cost and watch the gains turn to losses, to raise an external cost and see a market quietly overproduce — this is the book’s thesis under your own hand. The numbers are not fixed. They never were. Take hold of one and see.
The first price. Two curves describe a market: a demand that slopes down, because each buyer will pay less for the next unit, and a supply that slopes up, because each unit costs more to make than the last. Where they cross is the only price at which the quantity offered equals the quantity wanted — the market’s verdict. Shift either curve and watch the price chase the crossing. The shaded wedges are the gains from trade: buyers’ above the price, sellers’ below it.
Raise demand or cost and the equilibrium slides along; steepen sensitivity and the same shift moves price more, quantity less.
The price is nobody’s decision. It is what is left standing when what buyers will bear meets what sellers will accept — a truce, not a measurement.
The most important rule in economics, made into a dial. Each extra unit brings a marginal benefit (falling, as you tire of it) and a marginal cost (rising, as it grows harder to supply). The right amount to do is not as much as possible, nor as little — it is exactly the amount at which the benefit of the next unit equals its cost. Slide the chosen quantity and watch the total net gain. Push past the meeting-point and the curve turns down: you are now paying more for each unit than it returns.
The dashed line marks the optimum, where the two curves meet. Try to find it with the quantity dial; raise the cost and watch it move left.
Stop where the next unit’s benefit equals its cost, and not a step beyond. Almost every optimum in the book is this one sentence, applied to a different quantity.
What a seller does when no rival disciplines it. A firm with market power knows that selling one more unit forces down the price on all the others; its marginal revenue therefore falls faster than demand. It produces only up to where that marginal revenue meets marginal cost — less than a competitive market would, at a higher price. The faint point is what competition would have delivered; the bold point is the monopolist’s choice; and the red triangle is the surplus simply destroyed — trades that would have helped both sides, never made.
The lower the cost relative to demand, the larger the markup the monopolist can take — and the bigger the red triangle of lost trades.
Power shows up as a quantity withheld. The monopolist’s gain is smaller than the buyers’ loss; the difference, the red triangle, is no one’s — it is simply gone.
A cost the market cannot see. The factory weighs only its private cost; but each unit also imposes a cost on others — the smoke — so the social cost lies above. Left alone, the market produces where demand meets private cost, which is too much: every unit past the social optimum costs the world more than it is worth, and the red triangle measures the damage. Raise the external cost and watch the gap open. Then apply the corrective tax — exactly equal to the harm — and watch the market move itself to the right place.
With no tax, the market overproduces and the red triangle is the loss. The tax shifts private cost up to social cost; the triangle vanishes.
The market is not wrong so much as blind: it cannot price what it is not charged for. Make it pay the true cost, and it corrects itself.
What a promise of money later is worth now. A sum due in the future is worth less than the same sum today, because today’s could be lent out and grow; the interest rate is the price of that waiting. The curve shows how a future $1,000 shrinks as it is pushed further off — and how a higher rate bends the curve down harder. Move the rate and the horizon, and read the present value: the verdict the interest rate passes on time itself.
At a zero rate the curve is flat — time costs nothing. Raise the rate and the distant future is discounted nearly to nothing.
The future is not worth its face. To price time is to decide how much of tomorrow we are willing to carry into today — and the rate is that decision, written as a number.
Why a small difference in growth is the largest difference there is. Two economies begin identical at 100. One grows a little faster than the other, year upon year, and the gap between them does not add up — it multiplies. Over a lifetime, a single percentage point separates the comfortable from the rich. Set the two rates and the span of years, and watch them part. Then switch to the log scale, where steady growth becomes a straight line, and the faster economy is simply the steeper one.
One point of growth, sustained, is the difference between a country that doubles in a lifetime and one that barely moves.
Compounding is the quietest force in the book and the most powerful. Nothing else turns so small a difference, given only time, into so vast one.
Why people refuse a fair bet. The rational agent values a dollar gained and a dollar lost alike. Real minds do not: a loss is felt more keenly than an equal gain — and the value curve bends sharply at the origin, the kink. Set how much heavier losses weigh (loss aversion) and how quickly feeling dulls (curvature), and look at the two marked points: the joy of winning fifty against the pain of losing fifty. Then read the verdict on a coin-flip for fifty either way. Its true value is zero. Its felt value, for almost everyone, is a loss — so they walk away.
At a loss aversion of 1.00 the curve is symmetric and the fair bet is a toss-up. Above it, the kink appears and the bet is refused.
The bet is fair; the mind is not. Loss aversion is enough, all by itself, to make a rational gamble feel like a mistake — and the kink is where that feeling lives.
The contest in its purest form. Two players each choose to cooperate or defect; the table gives the payoffs, yours first. Whatever the other does, you score higher by defecting — so a rational player defects, the other reasons the same, and both land in the lower-right cell, the only stable outcome (marked). Yet both would have done better had they cooperated, top-left. That is the dilemma: individually rational, collectively ruinous. Set the moves and see. Then let the game repeat — where, met by a copycat, cooperation can finally pay.
In a single game, defection wins. Repeat it against a copycat and mutual cooperation becomes the rational course.
Self-interest, the engine that was supposed to serve the common good, here drives straight past it. The trap is no one’s error; it is the structure — which is why we build institutions to escape it.
Five more machines, reaching into the corners the first eight left dark: the choosing mind itself, the price it puts on risk, the failure that information breeds, the wedge a tax drives through a market, and the shape a whole society’s income takes. The rule does not change. Take hold of an input, and watch the verdict re-form.
How a chooser settles on a bundle. The straight line is the budget — every combination of two goods the income can just afford; its slope is the ratio of the prices. The curve is an indifference curve: bundles that please equally, bowed because variety is preferred to extremes. The best affordable bundle is where the budget line just grazes the highest such curve — the tangency, where the rate at which the chooser would trade the goods equals the rate at which the market makes them trade. Change a price and watch the line pivot and the choice slide along it.
Raise X’s price and the budget line swings in on the X axis; the chooser substitutes toward the good that has grown cheaper by comparison.
No utility is ever measured here; only compared. The chooser need not say how much a bundle is worth, only which of two they prefer — and that, the tangency shows, is enough to fix the demand.
Why a sure thing beats a fair gamble — and what the difference is worth. Because each extra dollar of wealth adds a little less satisfaction than the last, the utility curve bends. A coin-flip between more and less wealth therefore yields, on average, less satisfaction than holding its average for certain: the expected-utility point sits below the curve. The certainty equivalent is the sure wealth that would feel just as good as the gamble — and the gap between it and the gamble’s average is the risk premium: exactly what a person would pay for insurance. Bend the curve harder and watch the premium grow.
At zero aversion the curve is a straight line and the premium vanishes — a fair bet is a matter of indifference. The more it bends, the more certainty is worth.
The premium is the price of fear, and it is perfectly rational. Insurance is the business of selling certainty to the risk-averse for a little more than the risk is worth — and the gap on this chart is where that business lives.
How a market can destroy itself when one side knows more. Sellers know whether their used car is sound or a lemon; buyers cannot tell, so they will pay only what an average car is worth. But that average price may be too low to tempt the owners of sound cars, who withdraw — which lowers the average further, tempting still more good cars out, until only lemons remain. Slide the share of good cars and watch the buyers’ offer cross the line below which honest sellers cannot afford to sell. Past that point the market unravels: a failure not of price, but of knowledge.
When the good cars are few, the average is dragged down by the lemons, the honest sellers leave first, and the market collapses to lemons alone.
The cure is never price; it is knowledge — the warranty, the inspection, the reputation, the brand. Each is an answer to the lemon: a way to make quality visible, so the good cars can stay.
Who really pays a tax, and what it costs beyond its revenue. A tax of a fixed amount per unit drives a wedge between the price the buyer pays and the price the seller keeps; trade shrinks to where the gap fits. The shaded rectangle is the revenue collected; the red triangle is the deadweight loss — trades worth making that the wedge prevents, surplus that goes to no one. And who bears the burden is not set by who writes the cheque but by who can least escape: steepen demand, and the buyers, unable to walk away, absorb the larger share.
The steeper the demand — the harder buyers find it to do without — the more of the tax lands on them, whatever the law says.
A tax is two costs, not one: the revenue it moves, and the trades it kills. The first is a transfer; the second is pure loss — the price a society pays for raising the money at all.
How a whole society’s income is shared, drawn as a single line. Line up everyone from poorest to richest along the bottom; the curve shows the cumulative share of all income earned by the poorest such fraction. Perfect equality would be the diagonal — the poorest half earning half of everything. Real distributions sag below it, and the gap between the diagonal and the curve, doubled, is the Gini coefficient: zero when all share alike, one when a single household takes everything. Pull the dial and watch the curve bow, the gap widen, and the bottom half’s share fall away.
A single number stands in for a whole society’s distribution — which is its power, and its poverty: it cannot say who is poor, only how far apart the ends have drifted.
The Gini compresses a million incomes into one figure between zero and one. It is the most quoted number in the study of fairness, and the most treacherous: two very different societies can wear the very same coefficient.
Thirteen machines, and one lesson running through them. In every case the number the model reports — the price, the optimum, the markup, the loss, the present value, the gap between two economies, the felt value of a gamble, the payoff of a game, the best affordable bundle, the price of risk, the unravelling of a market, the burden of a tax, the reach of inequality — moved the instant you moved its inputs. None of it was given. Each was an outcome, waiting on a contest you could reach in and change.
This is the whole argument of the atlas, made operable. The volumes can tell you that a wage is a bargain and a price a truce and a growth rate a destiny; here you have felt the levers under them. The figures in the twenty-six volumes are these same models, drawn once and held still — and now you know that behind each frozen engraving is a mechanism that would have produced a different number under a different hand.
Step back to the atlas to choose a volume, read the preface for the argument entire, walk the guided tour for a path through the plates, consult the index to find any idea by name, open the constellation for the whole web of bonds, weigh the elements in the properties, combine them in the compounds, or follow one idea across the table in the diagonals, hear the case against the book in the objections, or set the ideas in time in the chronology, or find the contest inside the numbers in the measures, or set the table to work in the cases, or find where its reach ends in the priceless, or what it leaves out in the gaps, or how it is lived in the first person, or read through rival eyes in the schools, or the real works behind it in the sources, or the contests of our own moment in the unsettled. The instruments will be here when you return — and the numbers, as ever, will move when you ask them to. The argument remains open, because the contest never closed.