The Clock

Before the clock ruled work, the task did. The farmer worked while there was light and weather; the fisher went out with the tide; the smith stayed at the forge until the iron was worked. Labor took the shape of the job and the turn of the day and the season, and it stopped when the thing was done. Time, in such a world, was not a quantity a person spent — it was the medium the work moved through, with no separate existence to be saved or wasted. The anthropologist who studied the Nuer found that they kept their day by the round of their cattle, the milking and herding and watering, and had no word for time as a thing that passes, that one might hoard or throw away.1 In medieval England a length of time was named for the work that filled it: a span might run as long as it takes to cook an egg, or to say a prayer.2 The hour existed, but loosely, and almost no life was ruled by it.

What changed the world’s relation to time was a machine. The mechanical clock came first out of the monastery, where the bell that called the monks to prayer at the fixed hours imposed the earliest true schedule — a day cut into equal parts and obeyed by all. From the fourteenth century the clocks spread out of the cloister into the towns, onto the church towers and the market squares, and a new kind of time arrived with them: the merchant’s time, the historian Jacques Le Goff called it, set against the older time of the Church.3 This new time was unlike the task’s in every way that would come to matter. It was uniform — every hour the same length, whatever the season or the labor. It was divisible without end, into minutes and then seconds. And it was abstract, detached from any particular work — an empty vessel that could be filled with anything, or with nothing. Lewis Mumford would later judge the clock, not the steam-engine, to be the key machine of the industrial age, and he was right.4 The engine multiplied the work a man could do; the clock turned the time he did it in into something that could be measured — and so, in time, sold.

The factory turned the clock’s abstract time into a discipline. Where the old work had been paid by the piece or the task, the new work was bought by the hour — and an employer who has bought the hour has every reason to see that none of it runs to waste. So the factory acquired a clock, and a bell, and a locked gate; it fined the latecomer and the idler; it kept a record of each worker’s time and docked the pay of any who came up short. The day was no longer the worker’s to arrange around the work; it was the employer’s, measured out and accounted for. Thompson found the whole transformation gathered into a single distinction that the industrial worker came to feel and the old craftsman never had — the distinction between the employer’s time and one’s own.5 What the factory taught, in the end, was that time had a value, and that during the bought hours the value belonged to someone else.

The clock could measure time and the factory could command it, but neither could, alone, make a person feel that wasting it was wrong. That was the work of a moral campaign, preached from the pulpit and drilled in the schoolroom for two centuries. The Puritan and then the Methodist taught the sinfulness of idleness and the duty of redeeming the time; the charity schools beat punctuality into the children of the poor long before they reached the factory gate; and the maxim that pressed the whole teaching into four words was set down by Benjamin Franklin in 1748 — that time is money, and the hours a man lets slip are coins he has thrown into the street.6 It is a strange thought, held up to the light: that the minutes of a life are a currency, and rest a kind of spending. But it is now so deep in us that the other way of thinking is hard to reach at all — and it had to be taught, against the grain of every older way of living, before a bought hour could feel like a natural thing to sell.

The teaching did not take all at once. For generations the workers held to the older rhythms where they could — laboring hard as a deadline neared and easing after, and keeping ‘Saint Monday,’ the unofficial holiday of those who had spent Sunday too well to face the Monday morning.7 They resented the clock and the bell as the instruments of their masters, and for a time they fought the new time itself. But the deeper change was not the resistance; it was what the resistance became. The workers stopped trying to escape clock-time and began to bargain inside it. They accepted that the day would be counted in hours, and they organized to win some of those hours back — to set a limit on the working day, and to make every hour past the limit cost the employer more. The eight-hour movement was the proof of the conversion: it did not reject the clock; it fought over how many of the clock’s hours were the employer’s, and how many were the worker’s own.8

We are all the factory’s children now. Clock-time is so wholly inside us that we can scarcely feel it as anything but natural — we wake to it, are paid by it, measure our own worth in it, and forget that it was made. But made it was, in history, by a machine and a discipline and a long campaign of moralists, against a world that had got along without it through all the ages before. And nearly everything the law of the wage and the hour does stands on the thing they made. The minimum wage is a floor under the price of an hour. The overtime premium is a surcharge on the hour past the standard. Off-the-clock work is labor for which the bought hour went unpaid; wage theft is the taking of hours owed; and the entire apparatus of the timesheet exists to count the hours and settle who owns them. The field treats the hour as its given — the unit it adds up and litigates over — and never thinks to ask where the unit came from. It came from the clock. The clock made the hour into a thing, and in the making handed the field its subject, and half its name. But a unit is not as simple as it looks, and a subject taken for granted is a subject not yet understood. To see what the law is really standing on, the made hour has to be taken apart.

To say that the clock made the hour into a thing is true but not yet precise, because the hour is not a single abstraction. It is a stack of them, each a distinct achievement, each won against the grain of how time is actually lived. Four of them carry the weight of everything that follows.

The first is uniformity. The clock made every hour the same length — and this was less obvious than it sounds, for the hour had not always been uniform. The medieval day was kept in temporal hours: the span of daylight was divided into twelve and the span of darkness into twelve, so that a daytime hour in June ran far longer than one in December, and an hour of the day was never the same length as an hour of the night. The sundial and the water-clock tracked these unequal hours faithfully enough. It was the mechanical clock, beating at a constant rate, that could tell only equal hours — and as it spread, the equal hour became the hour, and the seasonal one was forgotten.9 Uniformity, the most natural-seeming property of the unit, was the instrument’s gift and not the world’s.

The second is detachment. The clock made time an empty vessel, severable from whatever fills it — so that an hour became a quantity with no necessary content, the same whether spent in hard labor or in idle waiting or in nothing at all. This is the reduction that permitted the hour, rather than the job, to be the thing one buys.

The third is divisibility. The clock cut time without end — into minutes, and then seconds, and then finer still — so that the unit had no smallest part, and any span of it, however brief, could in principle be named, marked off, and counted.

The fourth is commensuration. This was Franklin’s cut, and the moralists’: the uniform, detached, divisible unit was set equal to money, at a rate. Once an hour could be exchanged for a sum, time became a kind of currency, and the whole vocabulary of owing and earning and stealing fastened onto it as though it had always belonged there.

These four — uniform, detached, divisible, priced — are not a historian’s ornament laid over the subject. They are the parts the unit is made of. And the law of the wage and hour, which takes the unit as simply given, is in truth standing on all four at once: leaning on each for a different thing it needs, and quarreling, at the edge of each, over a different question it cannot finally close.

The deepest of the four is detachment, because of what it does to the thing being sold. When work was bought by the task, what changed hands was labor: a finished thing — a shirt sewn, a field reaped, a load hauled — and the bargain was complete when the thing was done. When work is bought by the hour, what changes hands is time; and time is not labor but the capacity for it. The employer who buys an hour has not bought a quantity of work. He has bought a power, which he must then set to work, and the amount of work he draws from it is fixed nowhere in the bargain — it is fixed by his command of the hour he now owns.

Marx gave the two things their names — labour-power, the capacity, which the worker sells; and labour, its exercise, which the employer directs — and one need not follow him a single step further into his theory of value to see that the distinction is the hinge of the entire field.10 For it opens a gap that the world of the task never had: the gap between the time bought and the labor got. In a piece-rate world there is no such gap; the pay is the product, and there is nothing to dispute but the count of products. In a time-rate world the pay is the hour, and what happens inside the hour is a separate matter altogether — a matter of discipline, of supervision, of how hard and how long and at what the bought power is made to work. The bargain sets the price of the hour. It does not set the contents of the hour. And the whole law of the wage and hour lives in that opened space, between the hour the worker sells and the labor the employer takes.

It follows that the unit the law counts is not, in the end, a unit of time at all, but a unit of subjection — an hour across which the worker’s capacity stands at another’s command.11 This is why the plain question — was the worker working? — so often has no answer on the face of the clock. The clock records presence; the law must decide subjection; and the two do not coincide. The worker who waits beside a silent machine, ready, is producing nothing — but his capacity is held. The worker held in line at the gate to be searched does nothing his employer can sell — but he is not free to leave. Time itself, the bare abstraction the clock handed over, cannot resolve these cases; it only measures the span. What must be decided, in each hard one, is whether across that span the power was held — and that is a question the instrument was never built to answer.

Why time, and not work? The first movement showed that the hour became the unit; the section above showed that to buy the hour is to buy a capacity, not a quantity of labor. But the deeper question is why the wage bargain settled on time at all — why one buys the hours of a person rather than a stated amount of work from them. The answer is that the work cannot be written down in advance. A contract that tried to name every task the worker would perform, in every contingency that might arise across a day or a year, would be impossibly long and still unfinished, because the future of the work is unknown at the moment the bargain is struck. Ronald Coase saw, in 1937, that this is the very reason the firm exists: where the market would demand a fresh bargain for each separate task, the firm substitutes a single standing bargain in which the worker agrees, within limits, to do as he is afterward directed.12

Herbert Simon gave the structure its exact form. The employment contract, he showed, is the purchase not of a specified performance but of authority — the worker agrees to accept the employer’s direction, within a bounded range, as to which tasks he will perform; the wage is the price of his entering that range.13 What changes hands is discretion over the worker’s effort, to be exercised later, as the need reveals itself. And the unit in which that discretion is sold is time. The hour is the natural vessel for authority precisely because authority is open-ended: one cannot buy ‘the work,’ for the work is not yet known; one can only buy the span of hours across which it may be commanded. The incompleteness of the bargain and the choice of time as its measure are a single fact seen from two sides.

It follows that the gap named a moment ago — between the time bought and the labor got — is not a flaw in the bargain but its very design. The contract is deliberately left open; it does not fix the contents of the hour, so that the employer may fill them by command as circumstances dictate. But contents left open are contents left unmeasured. The thing the law will most need to know — whether the worker’s capacity was in fact held to the employer’s purpose across this span — is exactly what the logic of the wage relation declines to write down. The hour is sold full of authority and empty of specification, and it falls to the law, afterward, to reconstruct what filled it.

Here lies the field’s deepest difficulty, and the root of nearly all its problems of proof. The quantity the law must price — held, subordinated time — cannot be directly observed. What can be observed is the clock: presence, punches, the timekeeping record. So the law is forced to reason from an observable proxy, clock-time, to an unobservable target, commanded labor-power — to meter an input it cannot see by the marks of one it can.14 Every hard case is a place where the proxy and the target come apart: the worker present but not held — the bona-fide meal, the genuinely relieved break — and the worker idle, or even absent, yet held nonetheless — the technician on call, the employee waiting under orders. The proxy is never the thing; it is only the thing that can be seen. This is why the burden of proof is built as it is: when the employer alone keeps the records of the proxy and has kept them badly, the law will not let the unobservability of the target defeat the claim, but lets the worker make the hours out by reasonable inference and shifts onto the employer the burden of the books he should have kept. Beneath the four reductions, then, lies a fifth fact, prior to them all — that the bargain which buys the hour is incomplete by design, purchasing authority over time in the place of a measured quantity of work, and leaving the real magnitude forever to be inferred from a proxy that never quite fits. The four reductions made the hour countable. Incompleteness is why there is anything to contest in the count — and the seams that follow are the places where the proxy tears, and the law must reach through the tear to the unobservable thing beneath.

Each of the four reductions, in making the unit usable, left a seam — a place where the abstraction parts from the reality it abstracted, and where, under enough pressure, the two can be pried apart again. The hardest and most heavily litigated questions in the law of the wage and hour are not scattered at random across the field. They sit, every one of them, at one of these four seams; and to see the field this way is to see that its doctrines are not a list to be memorized but an architecture to be understood.

Detachment’s seam is the edge of the day. Because time was severed from task, it became possible — and then necessary — to ask what, within the span a worker is present, counts as the work he was paid for. The question cannot even arise where labor is bought by the piece: no shirt, no pay. It arises only once the hour is the unit, and it has no answer the clock can give, so the law has spent the better part of a century supplying one by hand. Idle and on-call time turns on whether the worker is engaged to wait or merely waiting to be engaged — which is to say, on whether his capacity is held.15 The donning and doffing of gear integral and indispensable to the work is part of the compensable day, which runs unbroken from the first such act to the last, so that the legal workday need not begin at the moment the worker punches in.16 Time spent in a mandatory security line at the exit is, under federal law, a merely postliminary activity and not compensable at all.17 And when boundary claims of this kind first multiplied — after the Supreme Court let workers recover for time spent merely walking, inside the plant, toward their stations — Congress reached in and legislated the edge directly, passing the Portal-to-Portal Act to shear travel-to-station and other preliminary and postliminary time off the bought span.18 The edge of the day is nowhere given by the clock. It is drawn, and redrawn, by courts and by statute, in the slivers where the day’s filled center fades into its contested margins.

Divisibility’s seam is the bottom of the scale. The clock made time divisible without end, but the law flinches from following its own instrument all the way down. Under the federal de minimis doctrine, slivers of time too small or too irregular to be practical to record may simply be disregarded — ‘split-second absurdities,’ the Court once said, are not what the statute is for.19 But how small is too small is exactly the seam, and different sovereigns set the floor in different places. California has refused the doctrine outright, holding that an employer must pay for regularly worked minutes even when they are few, because the divisibility is real and the technology to capture it now exists.19 The same handful of minutes spent locking up after a shift is beneath the law’s notice in one jurisdiction and a compensable wrong in the next — not because the minutes differ, but because the two sovereigns cut the floor of divisibility at different depths.

Uniformity’s seam is rounding. If every minute is the same as every other, then a few minutes shaved here and added back there ought to wash out over time — and for a long while the law agreed: a neutral rounding policy, one that did not in the aggregate cheat the worker, was lawful.20 But the premise was always a fiction of convenience — minutes are interchangeable in the ledger, never in the life — and the fiction holds only so long as the true minutes cannot be seen. Once the timekeeping system records each worker’s actual punches to the minute, the convenience evaporates, and the courts have begun to say so: rounding of meal punches has been struck down, and where the exact minutes are captured, the demand is now for payment of all the time worked, the rounding fiction giving way to the recorded fact.20 Uniformity was a workable assumption in an age of paper. Electronic timekeeping is quietly withdrawing its license.

Commensuration’s seam is the regular rate. Before an hour can be priced, it must be settled what the price includes — and the law insists that the hour’s value is not merely the base wage but very nearly everything of value the work earns: nondiscretionary bonuses, shift premiums, and the like must be folded into the ‘regular rate’ before a minute of overtime is figured.21 The seam shows whenever a lump of money has to be poured back into the hours to find its rate. A flat bonus for completing a weekend shift is not earned by the hour at all, yet to compute the overtime it must be treated as if it were — and the law must then decide across how many hours to spread it. California spreads it across only the non-overtime hours actually worked, a choice that raises the hour’s reckoned value, and the premium owed, above what the federal method would yield.21 The commensuration the clock made possible — time set equal to money — turns out, at its seam, to be a standing question of exactly how the money is made to sit inside the time.

These are four questions, but they are one shape. In each, an abstraction that made the hour usable is pressed until it parts from the reality beneath it, and the law must choose whether to follow the abstraction or the fact. That choice is the daily work of the field. It is made, almost always, in the small — in slivers and spare minutes and the spreading of small sums — because the abstractions hold everywhere except at their edges, and it is only at the edges that anything is left to fight about. The figure below holds one ordinary day up to the light, and marks the seams.

There is a question hidden inside the made hour that the world of the task never had to ask: who says how long it is, who says what it is worth, and who certifies how much of it was spent. Under the task these were not questions at all. The work decided when it was done; the daylight decided how long there was to do it; no one owned the measure, because the measure was the world itself. To make the hour a thing was therefore also, and unavoidably, to make an authority over the thing — for an abstract unit, unlike a finished task, does not measure itself. Someone must hold the standard, keep the clock, and do the counting.

And at first that someone was the master. The factory clock hung on the master’s wall; the bell was rung on his order; and the workers, Thompson records, suspected — often rightly — that the master’s clock was not above being held back in the morning and hurried at night. To own the hour was to hold a small sovereignty over time, and the early factory held it whole: it set the length of the day, it fixed the price of the hour, and it kept the only count that counted.

Seen from here, the whole law of the wage and hour is the long expropriation of that sovereignty — the slow taking of the master’s three powers, one by one, and their lodging partly in the worker and partly in the state. The minimum wage takes the price: the worth of an hour is no longer wholly the master’s to set, because the law now sets a floor beneath it. The overtime premium and the standard day take the length: how long the ordinary day runs, and what the hours past it must cost, are no longer his alone to decree. And the law of records and proof takes the count — most pointedly in the old rule that where an employer has kept no adequate record of the hours, the worker may make them out by reasonable inference, and the burden falls on the employer to show otherwise.22 The count is no longer the master’s to control by the simple expedient of keeping no books. Price, length, count: the three sovereign powers over the unit, each prised loose and redistributed.

This is why the field is not what it appears to be. It appears to be administration — the dull, additive bookkeeping of a natural unit, hours summed and wages divided out. It is in fact adjudication — the continual settling of who may define, measure, and price a unit that nature never supplied and that someone, always, must have the authority to fix. The oldest question in the political economy of work is not, at bottom, what an hour is worth. It is who may say. The clock made the asking possible; the law of the wage and hour is the room in which the asking is, still, carried on.

One thing more must be said, because it turns the whole story around. The clock’s four reductions were forged for the master. Uniformity, detachment, divisibility, price — these were the abstractions that made labor measurable, and what is measurable is what can be commanded, compared, driven, and bought cheap. The instrument was built to meter the worker.

But an instrument that can count is an instrument anyone can count with. The same uniform, divisible, recorded hour that let the master meter the day let the worker, in time, prove it — and demand it paid. The timesheet is a blade that cuts both ways: the master’s tool of discipline and the worker’s tool of proof, the very same object, simply turned around. The eight-hour movement was the first turning, when workers stopped trying to escape the clock and began to wield it, fighting no longer against measured time but for a smaller measure of it. The wage-and-hour case is the latest. It takes the master’s own instrument — the records, the punches, the minutes the employer kept in order to control the worker — and turns them into the evidence that the worker was not paid for all he gave. The fourteen minutes a day carved off at the edges, beneath anyone’s notice one shift at a time, are recoverable precisely because the clock recorded them; multiplied across a year and a workforce, the slivers become the case.

So the field is not the neutral arithmetic it resembles. It is the eight-hour war, carried on by other means — no longer over the length of the standard day, which is largely won, but over the seams of the constructed hour, fought now in the small places where the abstractions wear thin. The hour was the master’s invention. The law of the hour is the long account of how it was taken back, minute by recorded minute, and made to answer to the one whose life it counts.