The second plate of this volume held that the wage equals the value of the last hour worked, and granted, almost in passing, that the premium skilled labour commands follows the marginal product of the skill. But it left the harder question standing: why should one worker's marginal product tower so far above another's — why is the surgeon's hour worth a hundred of the porter's who wheels her patients down the hall? This plate takes up that question with the orthodoxy's most powerful and most consequential answer. The surgeon earns more, it says, not by accident and not by mere power, but because she invested — years of schooling, the long forgone earnings of her training, the labour of acquiring a difficult craft — in raising her own productivity, and her wage is the return on that investment. It is the worth thread carried to its furthest reach: not simply that one is paid one's product, but that one's product is high because one built it. And it is, in its way, a radical and a democratic thought — the worker reconceived as a capitalist, owning and improving the one asset he can never be parted from, himself.
The idea has a precise birth. For a long time economists had shied from it, feeling that to speak of human beings as capital was to demean them, to set them beside the machines and the livestock on the asset side of the ledger. It was Theodore Schultz who, in a presidential address to the American economists in 1961, made the notion respectable, and Gary Becker who, three years later, gave it its rigorous theory, and Jacob Mincer who fitted it to the data. The logic is the logic of any investment. A young person who stays in school forgoes the wages he might have earned and pays the costs of his training, in the expectation of a stream of higher earnings across the working life to come; the decision can be appraised exactly as a firm appraises a new plant, by the rate of return it promises. Mincer's earnings equation — log wages rising, year by year, with schooling and with experience — became the workhorse of the field, and its central estimate, that an added year of schooling lifts earnings by something near a tenth, one of the most replicated numbers in economics. The framework swept labour economics, and development economics besides: the differences among wages became differences in human capital invested, and the wealth of nations was found to lie, in great part, not in their soil or their machines but in the trained capacities of their people.
Becker drew a distinction within the idea that has proved as durable as the idea itself: between human capital that is general and human capital that is specific. General skill — to read, to reason, to write a clear hand, the broad competence a degree attests — is valuable to any employer, and so, Becker reasoned, the worker must pay for it himself, by accepting lower wages while he learns; for an employer who footed the bill would only see the trained worker poached away by a rival able to offer more, now that the training was done and paid for by another. Skill specific to the single firm — the knowledge of its peculiar machines, its people, its ways — is valuable nowhere else, and so its cost and its return are shared between the worker and the firm, which binds the two together: it is why the wage so often rises with tenure, why long-served workers are dear and turnover among them low, why the firm grows reluctant to let a senior hand go even in a lean year. It is a beautiful piece of reasoning, and it explains a great deal about the shape of a working life — though one might notice, in the sharing of the specific return, that a bargain has quietly re-entered the room, and with it the question of who holds the better hand in dividing the gain.
Human capital became, in time, the orthodox explanation not only of why one worker's wage stands above another's but of why the whole structure of wages has shifted beneath our feet — of why inequality grew. The premium that a college education commands over a high-school one, roughly steady through the middle of the twentieth century, climbed steeply after about 1980. The reigning account is skill-biased technical change: the computer and its many kin are tools that make the skilled more productive and the routine unskilled simply unnecessary, and so they raise, year upon year, the relative demand for educated labour. Claudia Goldin and Lawrence Katz framed the matter, memorably, as a race between education and technology. When the supply of the educated runs ahead of technology's appetite for them — as it did when first the high school and then the college spread through the population — the premium falls and the wage structure narrows. When technology runs ahead of the supply — as it did late in the century, when the long rise in schooling slowed — the premium climbs and inequality widens. On this telling, the great widening of recent decades is no failure of fairness at all, but the market accurately pricing a skill grown scarce relative to the demand for it; and the remedy is neither protest nor redistribution but schooling, more of it, for more people.
On this telling, the widening was no injustice but the market pricing a scarcer skill — and the remedy was simply more education.
And here the plate must admit the idea that has shadowed human capital from its first day, and that no quantity of data has yet been able to drive off: that education may raise the wage without raising productivity in the least. Michael Spence set it down in 1973, and it is as elegant as it is unsettling. Suppose that abler workers find schooling less costly than the less able — that the degree comes quicker and easier to them. Then even if the schooling teaches nothing of any use whatever, the able will acquire more of it, simply because it costs them less; employers, unable to see ability directly but able to see the credential, will rationally pay more for the credential as a sign of the ability that came before it; and the educated will earn more — not because the school made them productive, but because it sorted them, and sent a signal of what they already were. The wage-and-schooling correlation, the very bedrock on which human-capital theory stands, fits this rival story exactly as well. And there is evidence the signal is no fancy: the so-called sheepskin effect, the jump in earnings that comes with the completing of a degree, over and above the steady return to the years of study it crowns — the diploma itself, the parchment, carrying a value that the learning alone cannot explain, and that signalling predicts.
The stakes of the ambiguity are not academic in the least. If schooling is investment, then to subsidise it broadly is to enrich a whole society. If it is largely a signal, then to subsidise it broadly is merely to inflate the credential — an arms race of degrees in which each must run faster only to hold his place, and the diploma that once secured a post now secures only the right to seek one. We do not, even now, know the proportions. The prevailing and probably correct view is that education does both at once — builds some genuine capacity, and signals some prior ability — but how much of each remains among the most consequential of the discipline's open questions, and the confident policy of "more schooling for all" rests on an answer we do not securely possess.
Two further doubts trouble the passage from skill to worth, even where we grant that schooling builds real capacity. The first is a matter of measurement: the educated earn more in part because they were abler, or better placed, before ever they enrolled, so that the return we credit to the schooling is shot through with the selection of who undertakes it. Ingenious natural experiments — twins who happen to differ in their years of study, whole cohorts kept longer at school by a change in the law, children who grew up nearer or farther from a college — have laboured to isolate the true causal return, and they find it real, but often smaller than the raw correlation would suggest. The second doubt runs deeper, and it is the same one that undid the moral version of the orthodoxy two plates ago. Grant that the return to skill is genuine and causal. The capacity to acquire the skill — the quickness of mind, the steady early schooling, the family that could carry the years of forgone earnings, the accident of a neighbourhood and the schools that came with it — is itself, in overwhelming part, unearned. To credit the surgeon's premium to her merit is to forget how much of that merit was a gift: of an aptitude she did not choose, of advantages she did not earn, of a starting place on the track she did nothing to deserve.
Human capital, pressed into a theory of desert, launders the luck of birth into the appearance of earned reward.
And some part of the return to skill is no return to skill at all, but a return to a barrier. The premium a physician earns includes, beyond the worth of her training, the rent created by holding down how many physicians there may be — the licensing board, the rationed residencies, the credential that fences the profession round and keeps the supply scarce and the price of entry high. Occupational licensing now reaches a startling share of all the work that is done, and wherever it binds it lifts the wages of those inside by the simple expedient of keeping others out — which is the economics not of skill but of scarcity deliberately maintained, the matter, in its pure form, of other plates in other volumes. The figure we read at the surface as "the return to education" silently blends three quite different things: the genuine productivity that training builds, the signal that the schooling sends, and the rent that the credential collects at the gate. The orthodox phrase folds all three into a single number, and invites us to read the whole of it as the just reward of merit.
Return, then, to the question the volume keeps. The Skill is the worth thread carried to its furthest and most persuasive point: wages differ, it holds, because productivities differ, and productivities differ because people have invested, in varying measure, in the building of them. And there is deep truth in it — schooling does enlarge real capacities, skill does raise real output, and much of the structure of wages, and much of the wealth of nations, is written in human capital and cannot be read any other way. But even here, at the orthodoxy's very strongest, the word worth will not hold still. The premium the skilled command is part true productivity, part mere signal, part the unearned luck of being able to acquire the skill at all, and part the rent of a guarded gate — and the market reports the blend as one clean figure and lets us mistake it for desert. As an account of the ceiling — of what lifts one worker's marginal product above another's — human capital is indispensable, and this plate has tried to grant it every inch of its power. As a verdict on what a person is owed, it fails exactly where the orthodoxy always fails: it takes the value of an hour, in a particular and largely inherited arrangement of the world, for the worth of the one who works it. The wage of the skilled, like every wage this volume examines, proves in the end to be less a measurement than a settlement — of productivity and signal, of luck and rent and power — and the question of how much of it is truly earned remains, as the table is fond of saying, open.