The Apparatus

Strip the competitive model to its load-bearing member and what remains is one number: the elasticity of labor supply that an individual firm faces — how freely workers flow toward it when it pays a little more, and away when it pays a little less. The clean picture sets that number to infinity. Every humane or cruel thing the model has ever been made to say descends from that one silent setting. So the first instrument does only this: it hands the setting over. Move it, and three things move at once — the share of the worker's product that reaches the worker, the geometry of the firm's own decision, and the distance the market sits from a competition it approaches but never reaches.

The arithmetic underneath is exact, not stylised. At the point a firm with power chooses, the fraction of marginal product it withholds is one divided by the elasticity plus one; the wage is the rest. Where the elasticity is four, the wage is four-fifths of the product; where it is infinite, nothing is withheld and the worker is paid precisely their product. The curve below is that identity, drawn whole — and the flat ceiling it strains toward is the competitive wage, which the real economy, on the measured numbers, does not come near.

Read the instrument twice and the parent plate's inversion becomes a fact about geometry rather than a claim about politics. Perfect competition is not the rule from which monopsony departs; it is the lone point at the right edge of the curve — labor infinitely mobile — toward which markets tend and never arrive. The measured band sits in the curved middle. On these numbers monopsony is not a pathology that visits the occasional market. It is the general weather, and the flat ceiling the rare cloudless day.

An objection waits here, and it is a good one. The single buyer of the textbook — the lone mill in a mill town — is nearly extinct. If monopsony power required that figure, the whole apparatus would be a museum piece. But it does not. Three quite different stories about why a firm faces an upward-sloping supply of labor — fewness, difference, and friction — arrive, by separate roads, at the very same first-order condition. The markdown is one divided by the elasticity plus one, whatever put the slope there. This is the quiet engineering that lets the power be measured without first winning the argument about its cause.

That is the force of the next plate. It is not an equilibrium but a derivation, drawn as a confluence: three mechanisms entering from above, each sufficient on its own, none necessary, all discharging into a single expression. Because the expression is the same, an empiricist can estimate the markdown from observed behaviour — chiefly from how readily workers quit when pay changes — and stay wholly agnostic about which mechanism, or which blend, produced it. The diagram earns its keep by making that agnosticism look like what it is: not evasion, but economy.

One implication deserves to be said plainly, because it overturns the usual intuition about scale. Power here is not a property of bigness. A market of many small employers, each offering a slightly different commute, schedule, and culture, can carry a markdown as deep as a market with three giants. The wedge is set by the slope of the supply each firm faces, and that slope can be manufactured by difference and friction in a crowded field as surely as by sheer fewness in an empty one. The single buyer was never the claim. It was only the clearest cartoon of a condition that turns out to be general.

Now the plate that matters most, because it touches the oldest belief in political economy: that equity and efficiency are bought only at each other's expense — that whatever is done for the worse-off must be paid for in lost output, so policy is forever choosing a point on a downward-sloping menu. Call it the iron law of the tradeoff. Under competition it holds. Under monopsony it does not, and the failure is not a quibble at the margin but a hole torn in the middle of the doctrine. A floor set in a market with power can raise output and raise labour's share in the same motion. It is not redistribution. It is the repair of a distortion — nearer to breaking a cartel than to taxing the comfortable.

The geometry says it more honestly than any sentence. Put efficiency — total surplus — on one axis and equity — labour's share — on the other. The conventional menu is the bowed frontier: to move along it toward more equity is to slide down in efficiency, and that is the tradeoff everyone was taught. But monopsony does not sit on the frontier. It sits inside it, southwest of the competitive point, poorer on both counts at once — output lost to suppressed hiring, share lost to the markdown. From an interior point the two goods are no longer rivals. A correctly set floor moves the economy up and to the right together, toward the frontier it never reached. That diagonal is the whole of the case, and it runs across the grain of the menu, not along it.

This is why the argument cannot be answered by reciting the cost of redistribution. The cost of redistribution is real, and it is a cost of moving along the frontier. But the monopsony floor is not on the frontier when it begins; it is the journey toward it. The deadweight that competition would never have created is sitting on the table, and the floor is the instrument that picks it up. One may still set the floor too high and walk off the far side of the frontier into the old tradeoff — the plates have been candid about that edge throughout. The point is narrower and harder to dislodge: over the interior range, there is no tradeoff to be made, because there is no rivalry between the goods to begin with.

Every plate so far has pivoted on a single quantity — the competitive wage, the height the arch crests at, the right edge the curve reaches. It has one inconvenient property. No one can see it. It is the wage a market would set if the firm's power were removed, and that counterfactual is never run. Both sides of every minimum-wage dispute are arguing, whether they say so or not, about the location of the same invisible point. The disagreement is not really about the shape of the arch. It is about where its summit sits — and because the summit is unobservable, the identical policy can be, in good faith, denounced as job-killing and defended as job-creating at once.

This plate makes the invisibility the subject rather than hiding it. The arch is drawn as a faint family of candidates, because its true height is unknown. Two schools place the crest differently — two beliefs, drawn as densities along the wage axis above. Choose a school; the believed arch sharpens. Then drag the floor. Watch the same vertical line fall on the rising flank of one school's arch and the falling flank of the other's. Nothing about the law has changed. Only the unseen point has moved, and with it the entire verdict.

The practical doctrine falls out of the picture without further argument. If the summit could be seen, policy would simply name it and stop. It cannot, so the responsible procedure is to raise the floor in increments and let the employment response report back: still climbing, and the crest is ahead; flattening, and it is near; falling, and it has been passed. The dispute between the schools is then settled not by theory but by the walk itself — which is why the essay insists the question will not be closed in the seminar room, and why the only safe step is a small one.

Gather the argument and it resolves into a single lever. Every cause of monopsony power is something that presses the firm-level elasticity down — toward the company town, toward a deeper markdown. Every remedy is something that lifts it back up — toward competition, toward a wage nearer the product. The four sources and their four answers are not a list to be memorised but a balance to be read. They act on one quantity from two sides. Where the elasticity comes to rest is the markdown, and the markdown is the share of the worker's product the worker does not receive.

The final plate is that balance, drawn literally. The elasticity stands as a vertical pivot. The forces that suppress it enter from the left and press down; the policies that restore it enter from the right and lift up. Each source faces the remedy built to undo it: fewness against antitrust, contractual lock-in against mobility law, secrecy against disclosure, search friction against the floor. The diagram is the essay's whole apparatus folded into one instrument — and it shows why the law's habit of arguing about who may decide, rather than about the slope itself, so often leaves the lever untouched.

One corollary follows so sharply from the rest that it deserves its own plate, because it inverts a policy most people regard as unambiguously pro-worker. Subsidise low earnings directly — top up the paycheque, as the earned-income credit does — and in a competitive market the worker keeps the gift. Under monopsony, the firm catches much of it. The subsidy makes workers willing to take the job for a lower posted wage, and a firm with power does exactly what its power permits: it posts the lower wage. The worker's total rises, but by far less than was spent. The difference does not vanish. It is captured.

The plate sets the two instruments side by side. On the left, a subsidy to the worker shifts the supply the firm faces and the firm slides down it — posted pay falls, employment rises, and most of the public money lands as employer margin rather than worker income. On the right, a floor set at the same target wage cannot be pocketed: it is a price the firm must pay, not an offer the worker will undercut. This is why the two are complements rather than substitutes. The floor is what stops the subsidy from leaking. Spend without it and you are partly subsidising the very power you meant to offset.

Six instruments, one quantity. The first hands over the elasticity and lets it propagate — into the wedge it opens and the equilibrium it bends. The second shows that the wedge does not depend on the story behind the slope, so it can be measured before the story is agreed. The third finds, in the interior, the one escape economics rarely offers: more equity and more efficiency in the same motion. The fourth admits that the point everything is judged against cannot be seen, and turns that admission into a rule for acting. The fifth lays the causes and their remedies on a single balance. The sixth shows what happens to a subsidy when the balance tips toward the firm. Set in sequence they are not six arguments but one, viewed from six sides.

What unites them is the refusal to treat the competitive wage as the natural state of things. It is a single point at the edge of the first plate, a summit no one can locate on the fourth, the top of the pivot on the fifth — always the limit, never the locale. The discipline spent a century measuring departures from a baseline it had quietly assumed was the rule. The instruments here are built on the opposite assumption: that the slope is general, the markdown ordinary, and the cloudless day rare. Whether that assumption is right in any given market is, properly, an empirical question — which is exactly why these were built to be set and watched rather than merely admired.