The volume opened with the lone worker's weakness. Smith's asymmetry, in the first plate, set the masters few and patient against the workman single and hungry: a master could wait a year on his accumulated stock, a workman scarcely a week, and the wage that fell out of so unequal a bargain recorded the inequality of patience as much as the value of the work. This plate is the weaker party's answer to that asymmetry — the oldest answer, and the most bitterly contested there has ever been: combination. One worker, withholding his labour, is replaced by morning and starved out by month's end. But if all withhold together — if the labour of the whole shop, the whole trade, can be stopped at a stroke — then the employer can no longer pick the workers off one by one, nor outwait any single man, for there is no longer a single man to outwait. The union is the institution built to make that conversion and to hold it: to turn the individual bargain, in which the worker is weak, into a collective one, in which the two sides stand more nearly matched. It does not abolish the bargain the first plate described. It alters who, at the table, can afford to wait.
The union's fundamental act is the collective withdrawal of labour — the strike, or the ever-present threat of one. Against a single worker the employer wields two whips, the whip of replacement and the whip of time; against the organised whole, both fall slack at once, for there is no ready replacement for everyone and no waiting out a stoppage that empties the whole floor. And where the employer's power was precisely the buyer-power the eighth plate anatomised — monopsony, the wage held down below the value of the work because the worker could not cheaply leave — the union is its natural and exact answer: two concentrations of power set against each other across the table, a bilateral monopoly, in which organised labour can drive the wage back up toward the value of the marginal hour, and beyond. Here the union, like the minimum wage under monopsony that the eighth plate described, can lift the wage without costing the jobs the competitive model would foretell — because it is not fouling a competitive market but offsetting a power that was already in the room. The union, one might say, is the worker's monopsony, turned to face the other way.
The larger idea has a name, and an author. John Kenneth Galbraith, surveying in 1952 an economy that looked nothing like the textbook's teeming crowd of small competitors but was ruled instead by a handful of great firms, argued that the check upon concentrated economic power, in the world as it actually runs, comes far less often from the competition the theory longs for than from the rising of opposing power on the other side of the market — what he called countervailing power. The great retailer arises to discipline the great manufacturer; and organised labour arises to face organised capital. On this account the union is not, as its critics paint it, a spanner flung into a smoothly turning competitive machine. It is a response to the absence of competition — a counterweight hauled onto the scale because the scale was already loaded. Where power has gathered on one side of the wage bargain, the union is simply power gathering on the other.
The union is not a spanner in a competitive machine. It is a counterweight hauled onto the scale because the scale was already loaded.
Turn to the evidence, and to its genuine two-sidedness, for the union is not a thing to be praised or damned whole. Members of unions earn more than comparable workers outside them — a premium long put somewhere between a tenth and a fifth, larger for the less-skilled than the more — and it is a premium that compresses the spread of wages, drawing the better-paid and the worse-paid closer together, within a firm and across an economy. But Richard Freeman and James Medoff, in the careful reckoning that gave their field its title, insisted that the union wears two faces, and that to see only one is to misjudge the whole creature. The first is the monopoly face: the cartel of labour-sellers raising the wage, with whatever costs to employment and to efficiency such a raising can bring. The second is the voice face. Borrowing Albert Hirschman's distinction between exit and voice, they observed that a worker unhappy in his place has, in the open market, only the recourse of exit — to quit and try elsewhere; the union gives him voice instead — a collective channel to air a grievance, to press for a remedy, to mend what is wrong short of walking out — and voice, by stemming the ceaseless churn of quitting and rehiring, by lifting morale and the willingness to learn the job deeply, can make a workplace more productive, not less. The union that raises the wage may, by its second face, also raise the product from which the wage is paid.
And like the bargain it answers, the union was never granted; it was wrung out of a long and often bloody struggle, and the law that now shelters it spent the better part of its history bent on its destruction. The first plate marked the Combination Acts, which had made the very fact of the workman's combining a crime; their repeal, in the eighteen-twenties, was no more than the first grudging inch, conceded and then half-snatched back. For a century after, the strike was met with the injunction, the imported strikebreaker, and not seldom the bayonet. The high-water mark came late, and came suddenly: in the United States the Wagner Act of 1935 guaranteed at last, in plain law, the right to organise and to bargain collectively, and within two decades the membership of unions rose to a third of all who worked. And with that rising came the great compression — the steep mid-century narrowing of the American wage structure, the years in which the distance between top and bottom shrank as it has neither before nor since — one of the most arresting facts in the economic history of the century. The counterweight, once the law would suffer it to be lifted, visibly moved the scale.
What rose has fallen. From a third of the American workforce at the century's middle, the membership of unions has sunk to about a tenth, and in the private sector to scarcely one worker in sixteen. And the fall has tracked, with a closeness difficult to wave away, the great widening of inequality and the loosening of the old, faithful tie between what labour produces and what the typical worker is paid — the very divergence that the volume's second plate named as its puzzle and its last plate takes up to measure. The careful studies, watching the wage distribution stretch as union density melted, lay a substantial share of the rising inequality of recent decades at the door of de-unionisation — by one much-cited reckoning, something near a third of the widening among men. The mechanism is the one this plate has traced, now running in reverse: as the counterweight was lifted from the worker's side of the table, the wage slipped down from whatever the union had held it to, and the gap between what labour made and what labour was paid yawned open into the chasm the final plate will survey. The union's rise built the great compression; its fall helped build the great divergence.
The union's rise built the great compression; its fall helped build the great divergence.
And yet the union's case must not be made larger than honesty will bear, for the counterweight carries costs of its own, and they fall, awkwardly, upon labour quite as much as upon capital. The wage a union wins for its members it may win partly at the expense of those outside the gate — the non-member crowded into a slacker unorganised market, the jobless outsider for whom no insider bargains, the young shut out of a trade the organised hold close. The insider-outsider theorists made just this the centre of their account of the stubborn unemployment that has dogged the strongly unionised economies: that the employed insiders bargain for their own wages, and not at all for the jobs of the excluded. And the monopoly face is no fiction. Where the market truly was competitive to begin with, a union lifting the wage above that level is a cartel like any other, with the costs to employment and to the efficient placing of labour that all cartels carry — the featherbed, the restrictive rule, the standing resistance to the better machine. The union's effect upon the general welfare, in sober truth, turns on the baseline against which it works: set against an employer's monopsony, it restores what power had taken, and may cost the economy nothing; set against genuine competition, it exacts a monopoly's toll. Both cases are real; which one obtains is a question for evidence, not for slogans. The public-sector union, moreover — now the larger part of all organised labour — bargains against no profit-disciplined employer at all, but against a diffuse and distracted public, a circumstance with troubles peculiar to itself. And even the link from de-unionisation to inequality, strong as the evidence for it has grown, must share the stage with the skill-biased forces the fourth plate set out; the honest verdict is that the thread of power and the thread of worth are both at work in the widening, and that neither, pulled alone, will account for it.
Return, a final time for this volume's plates in print, to the question they all serve. The union does not make the wage just, and it does not make it true. A unionised wage is no more the worker's marginal product than a monopsony wage is; it is, like every wage this volume has weighed, the settlement of a contest, fixed by the balance of power at the table. What the union alters is only that balance. It is the weaker party's attempt to even a bargain that began, as the first plate showed, lopsided — to gather, upon labour's side, a power to set against the power that capital had gathered on its own. And so the union does not escape the thesis of this volume; it confirms it, and it completes it. The five plates now standing in print trace a single argument from end to end. The wage is a bargain — so the first plate. The orthodox name that bargain's outcome the measure of worth — so the second; a worth that human capital deepens, and that signal and luck and rent confound — so the fourth. The buyer's power can press the wage below its supposed mark — so the eighth; and against that power the seller's organised power can press back — so this one, the ninth. Worth and power, the two answers the volume set out to weigh, both prove real, and both prove partial; and the wage is where the two are reconciled — not by any law of nature, but by a contest whose terms are made, and unmade, by institutions and by struggle. Seven plates of this volume are yet to be written, and the question stays, as it must, open. But the spine of the thing now stands, from the lone worker's weakness to labour's collective answer: the wage is, first to last, the register of who — at the table where its price is set — has gathered the power to wait.