Numéraire
Volume the Twenty-second · of the table of twenty-six
The Aggregates · Group V · № 22 · Sh

The Share

labor against capital

The share is the fourth and last of the aggregates, and it concerns not the size of the product, nor its fluctuation, nor the worth of the money that counts it, but its division — how the whole that a nation produces comes to be split between those who supplied the labor and those who owned the capital. Growth is the rise of the product, the cycle its wavering, the level the value of the measure; the share is who gets what. The volume opens with the orthodox answer, descended from the marginal-productivity theory of The Wage: that each factor earns what it adds, so the split is set by productivity, fixed by the production function, and naturally stable — the view Paul Douglas gave empirical form, finding labor's share near three-quarters and seemingly constant. Against it stands the answer from power: that the labor share is set by the degree of monopoly, the markup that market power permits (Kalecki); that it is no fixed number but the perpetually cycling front of a struggle between labor and capital (Goodwin); and that it has, since about 1980, fallen across the world — whether by the cheapening of machines (Karabarbounis and Neiman) or by the rising power of capital; and that the machine itself, looked at closely, neither dooms labor's share nor secures it but moves it both ways at once — displacing labor from the tasks it did even as new tasks reinstate it in others — so that the outcome hangs on a race the technology does not settle (Acemoglu and Restrepo). The flagship takes the long view, across two centuries, and finds the division to be neither natural nor fixed but political, shaped by institutions and policy, so that productivity sets a benchmark while power sets the outcome. And so the volume arrives at the question the author set it, the question that runs straight back to the first element of the whole table, The Wage: whether the split between labor and capital is set by productivity, or by power.

in print · № XXII · 12 of 12 plates
← The complete table of volumes

The argument of the volume

Begin where the orthodoxy begins, with the claim that the division of the product is a law of production. The Wage established, with Clark, that each factor of production earns its marginal product — the value of what one more unit of it adds; the Share asks whether that principle, applied to the whole economy, fixes the split between labor and capital. Paul Douglas, with the mathematician Charles Cobb, gave the answer its enduring form. Fit a nation's output to its labor and its capital through an aggregate production function, and under competition the exponents of that function turn out to equal the factor shares: the portion of income going to labor equals the weight of labor in production, and likewise for capital. A theorem of Euler guarantees that the shares so computed add up exactly to the whole product, with nothing left over, so the marginal-productivity theory of distribution is internally consistent. And Douglas found, empirically, that labor's share held near three-quarters, decade upon decade, just as the theory predicted. The division of the product appeared to be set by productivity, fixed by technology, and naturally stable — the productivity answer in its strongest form, orthodox for half a century. But it rests on assumptions: perfect competition, a particular elasticity of substitution, a stable production function, and a constancy that has, in our own time, given way.

Against the law of production stands the answer from power, and the volume turns to it. Michal Kalecki proposed that the labor share is governed not by marginal products but by the degree of monopoly — the markup over cost that a firm's market power lets it charge; the fatter the markup, the larger the profit share and the smaller labor's, so that two economies identical in technique divide their product differently if one is more monopolized. Richard Goodwin set the share in motion, modeling the labor share and the rate of employment as a predator-prey cycle: high employment emboldens labor and lifts its share, but the rising share squeezes the profits that drive accumulation, so employment falls, labor weakens, and the share recedes — distribution as a perpetual struggle with no resting place, bound to the very business cycle that runs through the aggregates. Then comes the great fact the theories must explain: since about 1980 the labor share has fallen across most of the world and most industries, refuting the old constancy. Loukas Karabarbounis and Brent Neiman traced the fall to the cheapening of capital goods, substituted for labor where substitution is easy — a technological account that returns the cause toward productivity. But the account is contested, for the capital whose share rose may be not machinery but housing, or intellectual property, or pure profit, the rent of market power; technology and power contend over the same fact, and the numbers alone do not decide. Daron Acemoglu and Pascual Restrepo sharpened the technological account by refusing to treat the machine as one force: production is a set of tasks, and automation hands some of them from labor to capital, displacing labor and lowering its share, even as new tasks arise in which labor holds the advantage, reinstating it and raising its share — so that the labor share is the outcome of a race between displacement and reinstatement, and the recent fall, in their reading, is a race labor has been losing, partly because invention has been aimed at automating old work rather than making new.

The flagship takes the long view and resolves the contest by rising above it. Thomas Piketty, assembling two centuries of data, showed that the division between labor and capital has swung widely across history — capital's share high in the nineteenth century, compressed by the wars and the high-tax decades that followed, rising again since 1980 — so that the mid-century constancy Douglas enshrined was not a law but an interlude. For the tendency of capital's share to rise he proposed a mechanism, that the return on capital exceeds the growth of the economy, so capital and its income outgrow the whole and wealth concentrates, absent the countervailing forces of fast growth, war, or taxation. And from this he drew the deepest claim: that the division is not fixed by any law of nature but shaped by institutions and politics — by taxes, by the rules of property and inheritance, by the strength of labor, by the concentration a society permits — so that the mid-century compression was policy and the recent widening was that policy undone. Thus the author's question is answered: the split is set by both, but differently. Productivity sets the frictionless benchmark, the split each factor would earn if paid its margin; power, working through institutions, sets the actual division and its movements, and because what has moved is power-governed, the weight falls on power. With the share, the four great motions of the aggregate are complete — growth, cycle, level, and division — and the question returns to where the whole table began, at The Wage: the single wage and the aggregate share are one question at two scales, and the thread of power runs unbroken between them.

The argument remains open. What follows is an attempt to fix its exact edges.

The twelve plates of this volume

  1. i The Functionin print Cobb and Douglas; the aggregate production function, whose exponents are the factor shares, so the split looks set by productivity and naturally constant. Read the plate →
  2. ii The Residualin print the theory of profit; what capital claims after wages are paid, and why what is left falls to the owner rather than to labor. Read the plate →
  3. iii The Degreein print Kalecki; the degree of monopoly, the labor share governed by the markup that market power permits rather than by what labor adds. Read the plate →
  4. iv The Pendulumin print Goodwin; the labor share and employment in a perpetual predator-prey cycle, distribution as a struggle that never comes to rest. Read the plate →
  5. v The Declinein print Karabarbounis and Neiman; the labor share has fallen across the world since about 1980 — by cheapening machines, or by the rising power of capital. Read the plate →
  6. vi The Decouplingin print the pay-productivity gap; why the typical paycheck stalled even as output per worker climbed, and where the missing gains went. Read the plate →
  7. vii The Robotin print automation and the substitution of capital for labor; whether the machine takes labor's share of the product or makes new work for it. Read the plate →
  8. viii The Divergencein print the capstone: capital tends to outgrow the economy, so across two centuries the split has moved with institutions and politics — productivity sets the benchmark, but power sets the outcome. Read the plate →
  9. ix The Superstarin print market concentration and the dominant firm; whether a handful of superstars has captured the share of income once going to labor. Read the plate →
  10. x The Collectivein print collective bargaining and the union; whether organized labor lifts its share of the product, or only redistributes it among workers. Read the plate →
  11. xi The Wealthin print from flow to stock; how the yearly division of income hardens, through saving and inheritance, into the concentration of capital itself. Read the plate →
  12. xii The Tidein print globalization and the world labor market; whether opening to trade and offshoring lowered the labor share of the rich economies. Read the plate →

In closing

This volume took up the fourth and last of the aggregates, the share — the division of the national product between those who labor and those who own. Its subject was the functional distribution of income, the split between wages and profits, the long fall of labor's share and the great divergence of pay from productivity; and the question the author set it was the one that runs straight back to the first element of the table, The Wage: whether the split between labor and capital is set by productivity, or by power.

The volume set the two answers against each other. Charles Cobb and Paul Douglas gave the productivity answer its form: the aggregate production function, whose exponents equal the factor shares under competition, so that the split appears fixed by technology and labor's share holds near three-quarters, a law of production. Michal Kalecki gave the answer from power: the labor share governed by the degree of monopoly, the markup that market power permits, so that rising concentration lowers labor's share with no change in productivity at all. Richard Goodwin set distribution in motion, the labor share and employment cycling against each other in a perpetual predator-prey struggle with no natural resting place. And Loukas Karabarbounis and Brent Neiman documented the great fact, that the labor share has fallen across the world since about 1980 — tracing it to the cheapening of capital, though whether the cause is technology or the rising power of capital remains contested. And Daron Acemoglu and Pascual Restrepo recast the machine's part as a contest of tasks — automation displacing labor from old work, new tasks reinstating it in new — so that labor's share is a race between the two, and its recent fall a race that labor has been losing.

Beneath the contest the flagship found the resolution. Thomas Piketty, taking the long view across two centuries, showed the division to be neither natural nor fixed but political, swinging with the institutions and the policies a society chooses — compressed in the mid-century by tax and union and regulation, widening again as those gave way while the return on capital outran the economy's growth. So the question is answered: the split is set by both, but differently — productivity sets the benchmark each factor would earn if paid its margin, while power, working through institutions, sets the actual division and its movements; and since what has moved is power-governed, the weight falls on power. With the share, the family of the aggregates is complete: growth the size of the product, the cycle its fluctuation, the level the worth of its measure, and the share its division — the four great motions of the whole economy, and the share their moral center, for it asks not how large the product but to whom it goes. The thread that began at The Wage, the price of a single hour of labor, runs unbroken to the division of a nation's entire product; and with the aggregates closed, four of the six families of the table now stand complete.

The division of the product between labor and capital is not, in the end, a law of nature but an arrangement that societies make: productivity sets the benchmark each factor would earn in a frictionless world, but the actual split, and its movements, are governed by power — by markups and bargaining and the cheapening of capital, and beneath them by the institutions and the politics that write the rules. And because the split hardens, year upon year, into the concentration of wealth, the question the share leaves open is the one the volume was set to ask, the question that runs back to The Wage: is the split between labor and capital set by productivity, or by power?