This is not the age of the small man
Reviewing John Kenneth Galbraith's "The New Industrial State" (1967).
The recent uproar over Geese—whether or not the band is a “psyop”—makes two things clear. First, people hate feeling like they’re being manipulated. To suggest that Geese fans didn’t discover the band “organically”—that they were played, to put it uncharitably—does not go over well. Second: The “psyop” is table stakes for the music industry. Lots of famous artists, and less famous ones, too, employ the services of promoters like Chaotic Good Projects to grow their fan base. This does not mean that Geese is fake or makes bad music; their collaboration with Chaotic Good Projects could just as easily mean that they’re really good and, therefore, worth all the publicity they can get.1
Here at Crow’s Nest, Ltd., we’re big believers in the fundamental obscurity of personal preference. If you think you know where your opinions and desires come from—well, no, you kinda don’t. Our preferences as consumers and as people with full lives, including any preference to listen to Cameron Winter’s crooning, are to some degree socially determined.2 There are institutions of all kinds—Spotify, concerts, zines, your group chat, the radio, music magazines, the record labels—that mediate how you’d discover and react to new artists and new songs. Without these mediating institutions for shaping your tastes, you don’t ever discover Geese, or ever get to like them. But it goes the other way, too: Without these mediating institutions, maybe Geese never goes big, and Cameron Winter is just a guy you see sulking around house shows.
In an economy where it’s possible to listen to many kinds of music, Geese has no reason to trust that anyone outside Brooklyn will find them and listen to them day-in, day-out. The everyday consumer is a fickle creature; if Geese needs listeners, those listeners need to be made to feel as if they want to hear Geese. As such, bands of all kinds will sign onto record labels and contract publicity agents and pepper Instagram with clips to maximize their reach, to ensure that consumers have no way not to discover them.
What this extended indie rock allegory is really getting at is the political economy of demand management. The existence of producer-supported mediating institutions that shape consumer preferences to support the success and longevity of those same producers—this is a model of the macroeconomy at odds with the conventional wisdom that the consumer is king. The simple Keynesian view that demand calls forth supply, that consumer preference and demand growth are what constrain and enable production, remains technically correct. But only technically.
The rest of the economy operates on this same principle, that producers have an interest in creating the consumer demand that props them up. Hence the relentless advertising—formal (Super Bowl commercials), semi-formal (something like a Costco deal), and informal (word of mouth, FOMO, Veblen-esque conspicuous consumption, etc.). American car companies must maximize the likelihood that you choose their gigantic, overengineered trucks, so their advertisements emphasize family, safety, comfort, and extreme off-roading. The producers of productivity software know you have no idea how to differentiate between their products—so they plaster ads across the New York City subway to make sure you know what your options are if questions of CRM workflows and Gantt charts ever come up at work. Discretionary appliances seem to be among the best examples of informal preference formation compounding over time: Nobody needs an air fryer—and I haven’t once seen an ad for one that I can remember—but they’re apparently only getting more popular. Pay too much attention and it feels like society itself is conspiring to shape your preferences in the interests of producers.3
The consumer, it turns out, is not king—and, emphatically, cannot be king. I understand this revelation may come as a blow to the egos of many Americans. But if we can stomach Geese paying for its publicity to ensure its longevity as a band, then it’s not a huge conceptual leap to imagine that the rest of the economy, too, is made up of producers strategizing how to capture and shape consumer demand. That goes for the producers of air fryers, Ford F-150s, and even nuclear weapons.
This fundamental likeness between the makers of kitchen appliances and the manufacturers of weapons of mass destruction is the lathe around which revolves John Kenneth Galbraith’s greatest contribution to economic theory, the New Industrial State (1967). It’s a surprisingly short tome which aims to lay out how large corporations of all kinds use various tools of supply and demand management and, more importantly, why: to capture and neutralize markets themselves, thereby eliminating uncertainty across large swaths of the industrialized sectors of the economy. But, as Galbraith cautions, corporations’ need to insulate themselves from markets means their capacity to advance the economy’s technological frontier must increasingly be underwritten by the state. This vision of the economy sees corporations becoming unacknowledged central planners serving at the pleasure of the state.
Galbraith is lauded as a public administrator—he ran the Office of Price Administration during World War II, handling price controls across the American economy, and was Kennedy’s ambassador to India—and is otherwise considered one of the leading exponents of Keynesian economic thought in the United States. His most famous book, the Affluent Society (1956), about the relationship between production and poverty and the importance of mediating institutions like civil society to change consumer preferences, was a smash hit at the time and presages the arguments in the New Industrial State. Despite his impeccable scholar-administrator credentials and his wonderfully acerbic prose, however, I rarely hear his work referenced in discussions of industrial policy today.4 This surprises me, because this book represents what I think is one of the most unique theoretical articulations of the relationship between technological innovation, the role of the corporation, market structure, and industrial policy that I’ve yet read. Galbraith presents a theory of the capital structure required for technological innovation that should undoubtedly inform the execution of industrial policy going forward.
On the cusp of this book’s sixtieth birthday, I want to take stock of Galbraith’s theory, and how its various facets hold up in a United States fretting about its lack of industrial capacity and staring down the barrel of climate change. How should we be reading this work of theory into today’s macroeconomic environment?
This review essay/reading guide proceeds as follows:
Technology and “the great producing organization.”
“Industrial planning is in unabashed alliance with size.”
How the 80s break the book.
Public enterprise and planning.
The growth models of “the last good generation.”
Innovation, industrial policy, and climate change.
Countervailing power.
Reading Galbraith into today’s political economy.
I. Technology and “the great producing organization.”
How is an air fryer like a nuclear bomb? Well, for one, they can both burn things to a crisp. But the similarities don’t end there. Both are produced through long-duration, capital-intensive, and organizationally complex industrial processes. They require expert labor and specialized capital inputs, sometimes so specialized that they cannot easily be reallocated to other uses.5 As such, timing matters: It would be unfortunate if, today, there was nobody to buy the air fryer or the warhead you had set aside the specific resources to manufacture last year. And both of these goods, while treated by many people as investments into their overall personal satisfaction and the stability of their home life, are technically unnecessary for human survival.
These particular qualities of advanced technology—economic complexity, production lags, and the lack of inherent necessity—are what spur corporations to the task of demand management. As Galbraith puts it, “The large commitment of capital and organization well in advance of result requires that there be foresight and also that all feasible steps be taken to insure that what is foreseen will transpire.” This is especially the case in a more affluent consumer market, where, at least among kitchen appliances, consumers can exercise ample choice over how precisely they would like to burn something to a crisp.6
The development of advanced technology and the management of end-user demand are, in Galbraith’s understanding, opposite sides of the same coin:
We have an economic system which, whatever its formal ideological billing, is in substantial part a planned economy. The initiative in deciding what is to be produced comes not from the sovereign consumer who, through the market, issues the instructions that bend the productive mechanism to his ultimate will. Rather it comes from the great producing organization which reaches forward to control markets that it is presumed to serve and, beyond, to bend the customer to its needs. And, in doing so, it deeply influences his values and beliefs―including not a few that will be mobilized in resistance to the present argument.
To that end, the large corporations in “the great producing organization,” must insulate themselves from markets—not just the market for discretionary demand, which they must manage to ensure adequate return on investment, but the markets for their inputs, too:
A firm cannot usefully foresee and schedule future action or prepare for contingencies if it does not know what its prices will be, what its sales will be, what its costs including labor and capital costs will be and what will be available at these costs. If the market is unreliable, it will not know these things. Hence it cannot plan. If, with advancing technology and associated specialization, the market becomes increasingly unreliable, industrial planning will become increasingly impossible unless the market also gives way to planning. Much of what the firm regards as planning consists in minimizing or getting rid of market influences.
Corporations making investment and production decisions in an increasingly technologically complex economy will prioritize reliable input supply and consumer demand over the abstract idea of a “free market”—which is really just occasion for uncertainties of all kinds. To that end, corporations will, when necessary, find ways to contract out of the market entirely: They can engineer the “reciprocal absorption of market uncertainty” through contracting or simply use vertical integration (in other words, buyouts) to achieve security of supply. Indeed, “the option of eliminating a market is an important source of power for controlling it.”7 Just as Thorstein Veblen’s experience of World War I informed his judgment that engineers were more valuable to the industrial system than businessmen, Galbraith similarly concludes that “the enemy of the market is not ideology but the engineer.”8
Galbraith thus neatly links the fate of an economy’s potential for technological development—product invention, innovation, and commercialization, for air fryers and nuclear reactors and everything in between—to its capacity to manage demand in the interests of its industrial “technostructure.” The development of a technologically sophisticated product requires the presence of a demand signal coherent and strong enough to outlast the capital expenditure cycle required to produce it, to say nothing of what it takes to cover for the associated uncertainties. A demand engineer, then, is just as important for a new technology as its product engineer.
II. “Industrial planning is in unabashed alliance with size.”
Galbraith has a bone to pick with the Brandeisian trust-busters who treat all large corporations and banks as manifestations of monopoly—and who see any opacity in the industrial economy as worthy of antitrust. “If big business and monopoly power tend to be identical, then all big business is inefficient and presumptively illegal.”9 But, in reality, “industrial planning is in unabashed alliance with size.” The development of new technologies and the investments in the capital goods that produce them only happen through the preservation of an industrial technostructure defined by its bigness—big enough to manage demand, big enough to plan supply, and big enough to invest into its own longevity under conditions of uncertainty.
Galbraith wants economists and policymakers to update their priors. If corporations in the industrial technostructure are competing for demand and hate uncertainty, they have an incentive to control not just their own prices but the quantities of products they sell. They will not salivate over starting a price war; doing so would only “de-range” the prices across their entire supply chain. They do not need to collude formally to reach equilibrium, either. And they would rather minimize losses than maximize profits, consistent with maximizing retained earnings sufficient to shield them from further uncertainty and to reinvest into longer-term asset-building. Galbraith quite eloquently postulates and proves each one of these claims to maintain that oligopoly is the price of an advanced industrial economy.
It’s not that small firms cannot exist in this milieu. But they must exist within a larger and industrially sophisticated technostructure that, through contracts and other kinds of balance sheet tie-ups, can insulate them from the uncertainties of the market. The aerospace and automotive sectors, for example, are chock-full of smaller contract manufacturers working almost entirely on behalf of the airplane and car manufacturing conglomerates. The prosperity of the latter underwrites the proliferation of the former, not the other way around.
Galbraith admits that the trust-busters are not wrong about their suspicions of private industry. But he insists that “size is the general servant of technology, not the special servant of profits. The small firm cannot be restored by breaking the power of the larger ones.” Insofar as Galbraith was the architect of FDR’s wartime price controls, he may have been the most qualified person in the country to grumble at his fellow progressives about their misunderstanding of the industrial structure. “This is not the age of the small man.”
Finance is not a major player in this model of the economy. Corporations in the industrial technostructure seek to optimize for retained earnings, the only kind of capital that they can use completely flexibly, rather than for equity or debt, both of which come with (sometimes quite costly) obligations. Galbraith quips that no banker would presume to know more than a General Motors engineer about the design of a new combustion engine—and the banker would lend anyway, because who would doubt a GM engineer, an “organization man”? The barrier to economic development in Galbraith’s time is no longer any scarcity of capital; rather, it is the concentration of expertise—which the industrial corporation is uniquely able to organize and shape to its ends.
There’s also not much of a role for organized labor. Not that it’s not important—Galbraith explains how widespread unionization and wage negotiations raise aggregate demand and allow the industrial technostructure to eliminate some uncertainties around its labor costs / wage bill—but the technostructure is inherently incentivized against deploying labor-intensive production methods whenever possible. “To substitute capital, in the form of machinery, the supply and cost of which are wholly or largely under control, for labor which is not and which can strike, is an admirable bargain. It is worth the sacrifice of some earnings. It is also adverse to the union for that is the purpose.” Unions are still necessary where labor-intensive production of goods and services remains unavoidable, including in many public-sector and service-provision vocations. But Galbraith remains a steadfast defender of Keynes’ “Economic possibilities for our grandchildren”: Freedom from toil is an inherent good.
This is not to say that industrial corporations do not care about their labor supply. Just as corporations must manage consumer demand, they must also manage their access to a supply of specialized workers with domain expertise: The labor of an industrial chemist, who could take their hard-won skills to any competitor, must be bought not just with money but with some mechanism of what Galbraith labels “identification” with the mission of their particular employer.
Ditch-digging is an unlikely passion for the average person. A useful completed excavation is a plausible goal of a group or organization. The problem is to win the surrender of individual preference in favor of the disciplined wielding of a spade. … If the ditch drains a particularly nauseous and malarial swamp, the individual on associating himself with the excavators may then become aware of the utility of their enterprise. This is to say that he finds the goals of the group superior to his own previous purposes and so he joins.
Pecuniary and psychic motivations can coexist. Galbraith gestures to the host of corporate tie-ups with university research labs and government initiatives to argue that corporations shape their workers into believing their vocations—whether air fryer or arms manufacturing—have some social purpose. By shaping the preferences of employees to ensure they identify at least somewhat with the mission of the corporation, the technostructure thus transforms its membership into “organization men.”10
The corollary to this argument, however, is that greater identification turns into a kind of psychic specialization: The farther you rise up the corporate ladder, the harder it is to define your identity outside your work. “For the corporation president,” Galbraith quips, retirement promises “only Stygian darkness.” It’s no wonder so many executives try to stay relevant after they retire.
III. How the 80s break the book.
The mature industrial corporation with a “technostructure” is a corporation that has cleanly separated legal ownership from formal control. Stockholders do not matter: Management treats them with the ritual reverence they require, pays them a dividend, and goes on with the operation of the technostructure. If individual stockholders want a greater dividend, they can sell their shares to express their displeasure without greatly affecting the corporation’s capital structure. Insofar as managers’ salaries are not tied to stock prices, Galbraith challenges his contemporaries to explain why a corporation will necessarily maximize its profits given its intrinsic incentive structure. His argument, which nicely skewers conventional wisdom, bookends one of his boldest quips:
The members of the technostructure do not get the profits that they maximize. They must eschew personal profit-making. Accordingly, if the traditional commitment to profit maximization is to be upheld, they must be willing to do for others, specifically the stockholders, what they are forbidden to do for themselves. It is on such grounds that the doctrine of maximization in the mature corporation now rests. It holds that the will to make profits is, like the will to sexual expression, a fundamental urge. But it holds that this urge operates not in the first person but in the third. It is detached from self and manifested on behalf of unknown, anonymous and powerless persons who do not have the slightest notion of whether their profits are, in fact, being maximized. In further analogy to sex, one must imagine that a man of vigorous, lusty, reassuringly heterosexual inclination eschews the lovely, available and even naked women by whom he is intimately surrounded in order to maximize the opportunities of other men whose existence he knows of only by hearsay. Such are the foundations of the maximization doctrine when there is full separation of power from reward … When capital was decisive and the capitalist was in control of the corporation, he maximized that which he provided, namely money. He specifically did so when his investment was small, when in consequence he had little to gain from improved earnings but when he could enrich himself greatly by looting the assets of the firm. The technostructure does not supply capital, but specialized talent and organization. There is, a priori, no reason to believe that it will maximize the return to capital. More plausibly it will maximize its success as an organization.
The technostructure is its own scarce commodity: a specialized organization.
Galbraith strongly contrasts the industrial technostructure and its aggregation of organizational capacity—without which it could not manage advanced technology—with its ostensible predecessor, the entrepreneurial corporation. Each has distinct technological needs and, by implication, distinct capital structures: Because the entrepreneurial corporation by definition has a less technically advanced production process, it needs less capital and needs to take on less long-term investment risks, and thus has less need to plan, and thus has no technostructure—or, at the very least, a much smaller one. Therefore,
This means, in practical terms, that if demand falls, [the entrepreneurial corporation] can adjust to it by laying off workers. The mature corporation by contrast cannot lay off its capital. The technostructure is large and costly and to curtail it is to disintegrate the very brain of the enterprise. … [Whereas] Control of the entrepreneurial corporation rests firmly on ownership. If it is not burdened by debt, it can ride out a temporary failure in earnings.
This difference in capital structure has psychic implications, too. Designed to funnel wealth to their owners, and with limited need to build an expert technostructure, entrepreneurial corporations are less easily able to shape feelings of social identification among their labor force: “In the entrepreneurial corporation men at all levels work, in principle at least, for the enrichment of someone else. This, as noted, is not an easy goal for men of average meanness to adopt.”
They are also inherently incentivized to resist regulation, particularly where higher wage and tax bills would cut directly into an entrepreneur’s take-home pay. Whereas industrial corporations can push tax bills and regulatory costs onto customers, onto shareholders, or onto their internal bureaucracies of lawyers and accountants—to weasel around them—it is not so for the entrepreneurial corporation:
By contrast, the entrepreneur paid for—and the smaller one struggled with—the regulation by himself. The burden of regulation like that of taxation is appreciably lessened by having it fall on someone else.
Since [the entrepreneur] sought, in principle, to maximize earnings, higher corporation taxes came out of that maximum and with incidence on himself.
This distinction between these two types of corporations—which will always be more of a spectrum than a binary—ends up being catastrophically important, because it’s the exact place where Galbraith’s entire theoretical framework collapses under the weight of events.
The Volcker Shock of 1981 raised corporate finance costs to crushing levels, “de-ranging” corporate debt service costs and forcing an industrial recession.11 The concomitant rise of private equity raiders, active stock management, and changes to corporate governance shifted the intrinsic incentive structures of industrial firms, too, to make them more resemble their entrepreneurial little siblings. Dividends and buybacks became key metrics by which financiers could judge the returns to investing in various corporations that, all of a sudden, were flooding Wall Street for exigent capital. At the same time, corporate executive compensation was made contingent on stock growth (and many employees got stock options, too). Industrial corporations were thus retooled into entrepreneurial corporations.
It’s hard not to read Galbraith wax lyrically on how the corporate system of his day could palliate its shareholders without sensing, immediately, how much has changed since he wrote it. His comment that a banker would never think to overrule a General Motors engineer has me writing “famous last words” in the margins. Ownership and control once truly belonged to the corporate technostructure; today they both belong to the asset manager descendants of the first private equity raiders. Shareholder primacy, once a ritual, has been made real.
Of course, the American industrial technostructure still exists. But Galbraith’s theory would suggest that the events of the 1980s, by putting private capital markets back in charge of the corporation, have left that technostructure far less insulated from the whims of short-duration investors and the whipsaws of market uncertainty than it ever was in the postwar decades.
And with less insulation from the market comes a greater incapacity for long-term capital planning. Galbraith’s arguments about automation seem to work in reverse to militate against industrial upgrading: Combine, on one hand, the de-ranging of capital costs and supply chain input costs for capital-intensive industries with, on the other, the great disciplining of organized labor and of workers’ bargaining power more broadly power over the last four decades—to say nothing of the rise of the gig economy—and the structural incentives for automation and for maximizing capital efficiency in labor-intensive sectors disappear.12 In Galbraith’s day, labor was a cost that needed to be managed while capital was a cost that could be managed; today, capital is a cost that needs to be managed while labor is a cost that can be managed. Thanks, Volcker. But take heart: Labor rights and capital efficiency are compatible!
IV. Public enterprise and planning.
Galbraith is mostly coy about his support for public enterprise in the New Industrial State because the book is predicated upon arguing that the distinction between public and private ownership doesn’t matter all that much; the technostructure is dependent upon the state regardless. But he devotes a chapter to listing the sectors that he thinks suffer from a lack of planning—not because they cannot be planned, but because various political and regulatory roadblocks have prevented the formation of a “national technostructure” that could successfully engage in capital planning and in the coordination and sequencing of investment. Galbraith explicitly endorses these sectors as ripe for public enterprise.
One of those sectors is transportation. Please forgive me for quoting Galbraith’s theory of the case at length, but I think this passage is one of his most lucid distillations of the book’s overall logic:
The clearest case [for planning] is urban and interurban surface transportation of people. This, it is clear in retrospect, required that there be one corporation, that is to say one planning instrument, covering the cities of an entire region including the lines between. The local systems would then have been developed in relation to the intercity and interregional system with joint use of rights-of-way, terminals and other facilities as appropriate. The prospective growth of the entire system would have been projected in a systematic and orderly way together with the investment requirements in the various parts and at various stages.
A planning unit of such scope and power would have been largely independent of local influences and pressures in setting fares. Prices, in other words, would have been wholly or largely under its planning control. It could have held its own with the automobile industry and the airlines in managing, i.e., promoting the demand for its services. It could have held its own with the automobile industry and the highway users in getting requisite public underwriting of its facilities—were costs and risks too great for it to carry, it could have pleaded military necessity as did the automobile industry and highway users in the case of the interstate highway system. Pleading further the doctrine of military necessity, it could have sought state underwriting of technical development. This would have placed it more nearly on a parity with the airlines which, in the last thirty years, have had many billions of dollars of subsidy in the form of military development of aircraft (ultimately usable as passenger vehicles) and in the development and installation of navigational facilities. The planning unit, assuming success, would have had internal sources of capital from earnings. This would have exempted it from petty interference by local governments or other sources of funds. It would have been able to make its own decisions on growth and technical innovation and would have tended to measure its success by its virtuosity in this regard. Its size and capacity for technical change, including automation, would have given it leverage in dealing with unions. Not least important, such a unit would have had a developed techno-structure in which group decision would have replaced the vagaries of individual competence.
None of this has happened. Local transit systems developed under public and private auspices and subject to local political influences and regulation. The railroads, under a different system of regulation, followed their own rather special pattern of development. Each part provided a fraction of the total services of moving people locally and regionally; none, in consequence, could plan the entire service. None had appreciable authority over prices, use of service, capital supply or labor supply. None had a developed technostructure. In an industry which required planning, none of the requisites of planned performance were available. It is not surprising that the results have been singularly bad.
Note how the concept of demand management retains pride of place in this explanation. A single coordinated technostructure can better manage labor relations, automation, government relations, system planning—and it can also shape consumer, industrial, and government demand to secure markets for its services well into the future. Galbraith has the advantage of being able to stand confidently aloof when necessary from the demands of labor unions and local governments—demands which many on the left might concede without questioning. He treats these entities primarily as political and market actors with their own prerogatives rather than as inviolable bastions of progressivism or democracy.13
Much of what counts as progressive economic thought today emphasizes the public sector’s need to coordinate the economy almost as a matter of course. Galbraith raises his eyebrows here, too: He suggests that this desired capacity to coordinate and discipline industry cannot simply be legislated or institutionalized into existence. It must be won through an entity empowered to shape the market around itself. The public corporation is the antecedent to public coordination.14
Were Galbraith to revisit argument today, he might be more openly in favor of public enterprise for one key reason: While the private sector has fallen victim to the curse of the 80s—insofar as shareholders now exercise ownership and control of much of the private industrial technostructure (often through closed-end mutual funds managed by private equity firms and asset managers)—the public sector-led corporation does not have to worry about shareholder management save for its statutory responsibilities to the federal government. Public corporations such as Amtrak, the TVA, and Fannie/Freddie all operate differently, with their own unique mandates as well as their own unique problems and roadblocks. But each is its own technostructure, and it’s hard not to argue that their freedom from shareholders—or, rather, their ritual deference to one single indefinite shareholder—makes them intrinsically more capable of longer-term planning in the interest of their own longevity. Within the public instrumentality, ownership and control remain separate.
One perennial problem that today’s average public enterprise faces, however, is a structural lack of liquidity. Retained earnings—the key to Galbraith’s flywheel of corporate capital planning—are simply not something that many public enterprises get much access to. (I have previously written about how some promising public enterprises simply can’t access cash as quickly as their private counterparts can.) Policymakers interested in rebuilding a public technostructure need to provide their public enterprises not just with adequate initial capitalization and capital market access, but with the capacity to finance the pursuit of their mandates creatively and flexibly.
Galbraith makes one more uniquely provocative intervention as to the use of public enterprise for industrial planning. Revisiting his time in India, where the Nehru-led government was, by the early 1960s, employing the tools of state socialism to set up public corporations of all kinds, Galbraith implies that the creation of an industrial technostructure is first and foremost a tool of capital formation and therefore shouldn’t be milked for jobs programs or subsidy schemes:
Social control [of corporations] bears most strongly on the two decisions which are of the greatest popular interest―on the prices charged the public and the wages paid to workers. This has the effect of keeping prices lower and wages higher than the most autocratic technostructure would permit. It eliminates net earnings and therewith this source of savings. The poor country, which most needs capital, is thus denied the source on which the rich countries most rely.
Another way to put it is that public corporations should not be molded into a kind of entrepreneurial corporation which deprives its own technostructure of access to retained earnings, funneling them instead toward propping up the public wage bill or the state’s budget.
Galbraith’s short detour into development policy is very clarifying when viewed in the bleary light of our much more boring contemporary arguments chewing on the mysterious essence of the “everything bagel.” As his logic would suggest, the primary function of the technostructure is to build up a capital stock and a planning apparatus that, when wedded, can weather uncertainty. Asking too much else of the technostructure cuts into the purpose of its capital structure.
V. The growth models of “the last good generation.”
Galbraith’s time at Cambridge—and the other Cambridge—put him in close contact with the leading lights of Keynesian economics, many of whom palpably shaped his thinking. I am taking this detour into the work of Galbraith’s peers to ground his applied work in a broader macroeconomic tradition. Viewed together, this Keynesian cohort’s contributions to macroeconomics reinforce Galbraith’s emphasis on oligopoly, demand, and planning.
Keynes argues that demand creates supply and that savings are mechanically produced by investment. It’s an insight that’s mathematically clean, more correct than its predecessors, and has lots of policy implications—chief among them being: Let it rip, brother. Rather than proceed from Keynes’ mathematics and abstract models, however, Galbraith takes the complexity of the modern economy as his starting point to argue that aggregate demand, above a certain level of subsistence, needs to be engineered through marketing, policy, government, or other means; in other words, the “great producing organization” that supplies demand must help create and channel that demand—particularly for advanced capital goods and consumer goods, few of which are prima facie necessary absent societal compulsion. In this way, Galbraith is also a fellow traveler with Joan Robinson, whose Theory of Imperfect Competition laid the foundations for treating oligopoly and monopsony as more characteristic of a complex capitalist economy than any notion of “pure competition.” Market structure sets the parameters for how demand ends up calling forth supply.
Michal Kalecki argued in “Political Aspects of Full Employment” (1943) that, while Keynes was correct about the benefits of pursuing full employment, capitalists prefer the power to fire workers over the gains that would accrue to them under a full employment-economy.15 Galbraith, in the New Industrial State, suggests indirectly that Kalecki’s assertion applies moreso outside the industrial sector—in entrepreneurial corporations and capital-light service industries, in particular. Galbraith stresses that the industrial technostructure, built as it is on specific arrangements of specialized expertise, will not as easily exercise the “power of the sack.” But he still affirms Kalecki’s work when he argues that capitalists are more likely to pursue automation and capital efficiency upgrades when the bargaining power of labor increases.
In 1955, Kalecki delivered a landmark lecture on economic development financing, arguing that shortages and bottlenecks in the “primary sector” of the economy—encompassing food, natural resources, energy, and perhaps key services like healthcare and housing—will constrain the role that demand-led growth can play in an economy. Inflation in primary goods prices and dislocation in primary goods supply chains will, by “de-ranging” prices across the consumer economy, dampen demand for consumer goods—a shock which feeds back into lackluster demand for capital goods. Here Galbraith and Kalecki connect: The stability of primary goods prices is essential for keeping the economy’s demand channels clear and for preserving investment pathways for long-duration capital goods production.16
Nicholas Kaldor, in his 1966 lecture on the “Causes of the Slow Rate of Economic Growth of the United Kingdom,” gets at this same argument from the opposite direction. His lecture can be abstracted to argue that effective demand sets the pace of the economy’s potential to build out a stock of capital goods and the technologically intensive processes for managing them. It’s easy to see how this argument supports Galbraith—who wrote part of the Affluent Society in Kaldor’s library.
Galbraith and Joseph Schumpeter, at first glance, seem to share a focus on how macroeconomic conditions constrain or enable economic complexity and technological innovation. But, in Galbraith’s own words, “I tried to struggle with the reality, however dull, and Schumpeter sought the spectacular, however irrelevant. … [He] would not discuss price theory and policy at all. I don’t think he ever made anything, so to speak, of imperfect competition.” This is not to say that Schumpeter’s work is irrelevant or doesn’t speak to Galbraith’s; his notion of “creative destruction” and his investigation into the political economy that powers it very well birthed the study of innovation economics. But Schumpeter’s emphasis on entrepreneurship, market entry, and the depreciation of the existing capital stock as the enabling conditions for innovation and value creation contrasts starkly with Galbraith’s insistence that innovation is in fact sped along by a stable and growing technostructure—not to mention Galbraith’s dismissiveness toward the innovative potential of entrepreneurs operating independent of a technostructure.
Galbraith gave credit to a lot of these particular economists for forming and challenging his thinking. In a 2002 interview, he mentions lunching with Joan Robinson in “the late 1940s or [early] 1950s” in Cambridge:
I met Joan and we went across the river and sat down for lunch and I opened the conversation by saying, “Joan, I’ve been out of touch due to the war, I’ve been out of touch with my colleagues in this profession. Who are the good people of the new generation?” And Joan’s answer was, “Ken, we were the last good generation.” She was very firm on that point.
I am undoubtedly missing many economists in this milieu.17 But this is just a selection of the star-studded cast of macroeconomic theory savants around whom Galbraith built his practicable industrial economics. The resonances between them lend considerable heft to the New Industrial State which, to its author’s immense credit, can riff eloquently off this broader “theory scene” without resorting to complex modeling.
VI. Innovation, industrial policy, and climate change.
Galbraith also said in his 2002 interview: “I never had the expectation of technology being the whole basis of entrepreneurship and its disastrous effect.” This was in the aftermath of the dotcom bubble, of course, and within the broader context of the rise of software and tech venture capital.
The New Industrial State, as we have seen, is about technology as much as it’s about macroeconomics. In an industrial economy, the two become inseparable. Galbraith argues that innovation cannot happen without comprehensive planning of some kind that insulates technologically complex and capital-intensive sectors from the vagaries of the market. He suggests that large incumbent industrial corporations will inevitably be first in line to commercialize new technologies—not just because innovation under conditions of oligopoly helps them preserve market share and capture new markets, but because they have the capital on hand to “learn by doing.” He assumes, by corollary, that small firms are not often the engine of new innovations; while this is not precisely true—small firms and contract labs have always been an important aspect of American innovation culture, in particular—Galbraith’s implication that innovations cannot be commercialized absent an industrial structure that can “bend the customer to its needs” at scale is very convincing.
The problem facing today’s policymaking classes hell-bent on revitalizing American innovation, then, starts with precisely what Galbraith was worried about: Technology itself cannot be the basis of entrepreneurship. A startup-heavy innovation landscape disconnected from channels of effective demand in the economy is not going to see its products commercialized. This is not a new problem, by any means, but it suggests that policymaking—particularly for a task like decarbonization—cannot merely focus on cajoling startups to commercialize complex new products themselves. Policymakers will instead have to shape consumer and industrial demand to stabilize the investment pathways for critical new technologies. Those rails connecting demand to supply will inevitably run through a technostructure.
One obstacle to translating lessons from sectors to the challenges of decarbonization is, it seems to me, the fact that only a few clean technologies and climate adaptation investments—solar, batteries, EVs, and home resilience investments—ever touch end-users directly. Much of the rest—nuclear energy, grid-enhancing technologies, and the like—will only be purchased by technostructures themselves as inputs to providing a less differentiable service to end-user consumers: electrons. It’s the same for low-carbon steel and concrete. In sectors like these, policymakers cannot rely much on skin-deep appeals to intrinsic product quality. They have to subcutaneously reshape existing and entrenched technostructures such that those organizations cannot envision their own longevity absent undertaking investments in these preferred pathways for producing clean materials or deploying clean energy projects. Or, of course, policymakers can create new technostructures where none existed before. But this requires capital, time, and the same questions of demand management. And leaning hard into demand-side supports such as advance market commitments or direct federal purchases does not, in and of itself, represent a long-term deployment strategy that outlasts the required capital outlays and informs iterative process innovations.
Another potential obstacle is one that lies just outside Galbraith’s model: The national economies of the 60s, where corporations play second fiddle to their domiciles, have given way to a globalized economy of multinational corporations. In a world of globalized demand for decarbonization solutions—a market where American demand, once preeminent, may be trailing off—can corporations still act as the submerged central planning arms of any one state the way Galbraith once imagined? Or does the fragmenting nature of global demand give multinational corporations the upper hand among competing states? To be sure, state-capitalist governance models across much of the world have never had the luxury of ignoring the global economy; many of the new state-owned enterprises emerging from the latest wave of nationalizations seek to employ global financial and trade networks rather than reject them. And corporations have often been disciplined into serving state policy goals even while chasing (and sometimes shaping) demand elsewhere. So I’m cautiously optimistic about the role of the state. But I wonder if a more coherent and less fragmented global demand landscape for decarbonization in particular—in other words, some supranational demand pull with emissions regulations and investment authorities, too—could help cement a more globalized technostructure focused on the thornier problems of industrial decarbonization, carbon management, and adaptation.
There’s a lot more I could say about innovation, the global economy, and climate change specifically. But insofar as new clean energy technologies and climate interventions are not intrinsically different from any other consumer or capital good that needs commercializing, what can we learn from past examples of how firms and governments shape demand and create a durable long-term deployment strategy? Maybe we need a “Got Milk?” campaign for clean energy; maybe, in the United States, we just need Democrats to lock Republicans out of power.18 Either way, we cannot rely on short-term price signals and the ostensibly low per-unit costs of clean energy to drive change for us. Total decarbonization does not proceed from good economics alone. True enough, clean energy needs a lot more capital. But capital alone is insufficient. Decarbonization needs an empowered technostructure—a globalized one.
VII. Countervailing power.
One of Galbraith’s most notable contributions to political economy was the notion of “countervailing power,” the suggestion that a growing civil society presents a beachhead from which everyday people can attack the influence of corporations on their lives and reshape demand itself in the interest of the public good. His preferred institutions of countervailing power are public universities, which produce workers with industrial domain expertise somewhat removed from direct identification with the social and political goals of the industrial technostructure. The importance of the industrial technostructure to Galbraith’s model of the economy is also why he saw countervailing power as so necessary. Only through the public contestation of “the good life” does the economy change in a more sustainable direction.
Galbraith introduced the concept of countervailing power in 1952 but expands considerably upon it in the New Industrial State’s discussion of the military-industrial complex and America’s consumption fetish. He’s no fan of the economy he’s describing, which prefers nuclear armaments to poverty assistance, and where corporations pretend that consumers have free choice but, of course, shape collective preferences through advertising. Just because “the great producing organization” shapes demand does not mean that it should do so unfettered by such considerations as public health and well-being.19 Perhaps that’s why I gravitated to his work so quickly: His model of the industrial economy—centered on how and why suppliers must shape demand for advanced products—is both accurate and justifies his political critique of it.
In our present era, where the United States government has dropped all pretense of caring about climate change and the challenges of urgent decarbonization, it is easy to see why countervailing power might be necessary. Civil society support and broad public narrative-building for climate change mitigation are what made the Green New Deal and Biden’s climate investments feel so urgent before they passed, after all, and it’s hard to believe that continuing this civil society-led effort wouldn’t still be necessary today if we care about delivering the most comprehensive response possible to the climate challenge.
That is one reason why Trump’s attacks on the various institutions that could amass countervailing power or can help build it among everyday people—the independent bureaucracy, the judiciary, the National Science Foundation, the National Institute of Health, the public arts, public universities, scientific and regulatory agencies like NOAA and the EPA, labor unions, and media—are so disheartening. These attacks leave our preference formation—including our preferences over the kind of political economy we are willing to tolerate—at the mercy of the likes of oil and gas companies, Silicon Valley grift startups, outspoken NIMBYs defending their property values, and the “folk wisdom” of social media algorithms.
Galbraith juxtaposes countervailing power against technostructure—but they are not really so different! Both are ultimately organizations of experts trying to shape consumer and industrial preferences, where the countervailing one seeks to level the playing field against “the great producing organization.” Applying this judgment to climate politics feels anodyne, but I’ll do it anyway: Nobody would disagree that decarbonization requires a large and empowered technostructure—but getting there requires building a robust politics of countervailing power. Those who would peddle in “climate policy denial” despite professing to care about decarbonization misunderstand this at the public’s and the planet’s peril.
It would be delightful if we could redirect the trillions of dollars we spend on war20 toward meaningful social and planetary security. Institutions of countervailing power are required to bring about that great reallocation. A large share of our armaments bill no doubt pays for bloodshed and bloat—but those trillions also reflect the sheer size and complexity of the military supply chain and carry immense economic multiplier effects, globally. It is not for nothing that some of our most innovative and complex industries—aviation, computer technology, and nuclear engineering, for example—are tied at the hip to defense procurement.
Galbraith, no friend of the Beltway Bandits, respects the problem enough to worry, presciently, that it might be impossible to safely zero out the trillions in our military budget in the interest of spending on such good things as healthcare spending, economic development, and environmental cleanup. These ends all seem so technologically non-complex, at least comparatively, that they would cost a fraction of those trillions—which sounds great unless you’re a Keynesian: Such a reallocation is, mechanically, a negative shock to public spending (and its investment multiplier) and would likely drive an immense economic contraction. Perhaps Galbraith is being too pessimistic; the good news about decarbonization-focused industries is that their production processes can be very technologically complex, and are therefore much better candidates for a great reallocation. But policymakers still need to make it past the “traverse,” the challenge of transitioning from one distinct and specialized capital stock to another without burdening the economy with mass unemployment or hyperinflation. Economists like Adolph Lowe, Wassily Leontief, and my friends and colleagues Melanie Brusseler and Yakov Feygin have all argued that state planning functions are necessary for even making sense of this challenge.
Institutions of countervailing power help ensure that the longevity of a technostructure does not come at the cost of the public good. But those same institutions of countervailing power require the perpetuation of technostructures to effect the dynamic transitions between capital regimes they might seek.
VIII. Reading Galbraith into today.
Even as Galbraith and his fellow travelers at both Cambridges laid the foundations for post-Keynesian and institutionalist strains of economic thought, Galbraith’s applied theory never seemed to become a school of its own. In fact, I think I have seen more of his work show up in economic history than in the history of economic thought.21 Now having read the New Industrial State, however, I think this omission is really unfortunate. Galbraith himself makes clear that most economists would do well to understand the economy as it actually exists and to delineate its principles deductively, rather than rely on inductively constructed abstract models. I can’t fault him for this: Galbraith is, fundamentally, an administrator, whose most important job involved understanding the economy at the level of its firms and their managers. His work is not just supremely well-written, and funny to boot, but it successfully balances his embrace of positivism, in that the economy can be measured and understood, with his appreciation for organizational and consumer psychology. And, thankfully, he wears his politics on his sleeve.
His optimism that the economy can be shaped in our interests is infectious, at least to me. It’s too bad that stagflation breaks his theory—but I think it breaks his theory in such interesting ways that his analysis of the industrial technostructure becomes more revealing, not less. Six decades on, his theoretical innovations and vocabulary can still be applied fruitfully to questions of macroeconomic planning, even in a fragmented global economy where the cost of capital disciplines the creation and maintenance of technostructures.
I could not help but read this book with the challenges of decarbonization and clean technology innovation in mind. For the past few months, I have been spelunking through various articles and books on innovation finance to see if there are theoretical approaches that policymakers’ “common knowledge” currently neglects. But Galbraith’s work is one of the few I’ve seen that postulates what kind of capital structure a firm needs to contribute to the uncertain task of innovation: one backed by retained earnings, the intentional and somewhat-hidden shaping of long-term demand, divestment of control from profit-maximizing shareholders, and, perhaps most importantly, bigness to shape markets around itself. Much of the rest of this literature, descended from Schumpeter, ends up focusing on the exertions of committed entrepreneurs somewhat to the neglect of both capital structure and how expertise can accrete around a corporate technostructure over time. The tension between Galbraith and Schumpeter lives on in the disagreement over the merits of entrepreneurship.
This disagreement can only be resolved in Galbraith’s favor. Returning to Geese frontman Cameron Winter, our serendipitous poster child for the political economy of demand management—there is no doubt he exerts himself creatively, or that many people consider his songs earworms. But is it easier to believe that his success comes purely from his direct efforts to build relationships with his listeners, who somehow knew ex ante that they sought his style of music, or through an interlocking and well-capitalized set of music institutions that amplify his work and give listeners and content creators reason to consider Geese relevant? This is how small acts go big. We should temper our dismay that impersonal forces—whether it’s the psyop of capital, of the party, or even of peer pressure—shape our very personal preference formation. Reading Galbraith offers a more radically optimistic promise: Anything is possible with organization.
To tell you the truth, I don’t like Geese. Sorry. Musically, I can vibe with it—the guitars are sick—but I just cannot stand the kinda-singing-kinda-crooning. It breaks my brain to imagine a full stadium of people crooning along to “Taxes” or something.
Whether you love it or hate it, you will know “Au Pays du Cocaine.”
I could say the same thing about the battery-powered Ninja blender I got from Costco, which, in hindsight, is one of the sillier purchases I’ve made this past year. I assume some product designer must have believed that a battery-powered blender reduces the space it takes up and makes it transportable. The truth is: It just makes for a more annoying blender. What’s the point of making a smoothie if the battery cuts out halfway through blending and you’ve got to charge it before finishing? For those readers wondering why I own this item… well, among other reasons, it was on sale at Costco over Christmas. And I thought it would be nice to have more fruit in my diet. Et voila.
Adam Tooze’s podcast about Galbraith, part of his series on heterodox economic thinkers, is a notable exception. And Zachary Carter’s Price of Peace is another one, insofar as the book treats Galbraith as among the preeminent American Keynesians. But Galbraith himself acknowledged that he never bothered to build a “school” around his work, unlike many of his peers at Harvard—such as Schumpeter, who built cohorts of disciples—preferring instead to remain a wide-ranging practitioner and writer:
I devoted a lot of time to my lectures, which tended to be also what I was writing about. They were, I think, safe to say, rather well attended. I did not spend a great deal of my time tutoring disciples. I saw a lot of them, but I never saw my main function as having a cadre of students taking my ideas to Washington. That was not a clear decision, it was the way things work out. … I couldn't have written the Affluent Society if I spent ten hours a day teaching, tutoring, and cultivating. (Dunn, 2002)
“It must not be supposed that group decision [and organized planning] is important only in such evident instances as nuclear technology or space mechanics. Simple products are made and packaged by sophisticated processes. And the most massive programs of market control, together with the most specialized marketing talent, are used on behalf of soap, detergents, cigarettes, aspirin, packaged cereals and gasoline.”
Galbraith’s logic here is that, once people have spent on their absolute necessities—food and shelter, primarily—they have the freedom to use their remaining savings however they’d like. The richer someone is, the more savings they accumulate relative to their basic necessities and, thus, the more discretion they can exercise. As such, corporations are invested in ensuring that you spend your savings on their goods. You will buy an air fryer.
Cue Dune: “The power to destroy a thing is the absolute control over it.”
Veblen’s “Soviet of Engineers” is updated as Galbraith’s shrug that both the Soviets and the Americans use the state to underwrite industrial planning and technological advancement. “Technological compulsions, and not ideology or political wile, will require the firm to seek the help and protection of the state.”
It is better, he argues, to separate measurements of absolute and relative firm size before calling the FTC.
“Identification can be increased by substituting automated for manual processes. This, at one step, reduces the number of workers susceptible to union goals and, by adding to the pay, interest and physical ease of those remaining, increases the tendency to identification.”
Inflation was, to be clear, going crazy at the time. There are many reasons, including but not limited to the 70s oil supply shocks, the demand-pull of Vietnam War arms procurement, upward wage renegotiations in response, and asset price inflation. Galbraith explains how and why the industrial technostructure was uniquely tolerant of inflation. But the rest of the economy clearly wasn’t. We will revisit this note in footnote 16.
I don’t know if I’m 100% right about this claim—it just feels like a provocative but nonetheless appropriate way to read Galbraith’s claims into the present. There are so many “innovative” industries in the United States that refuse to build up their own capital stock simply because they’ve found a business model that rewards them for avoiding that task. Uber, for example! But I think the jury is still out; we’re yet to see how capital-intensive variants like Waymo fare.
Galbraith’s style resembles Veblen in many ways—but the two come closest in their willingness to treat small-d democratic planning units, like labor unions and local governments, as, first and foremost, market actors shaped by incentives and, only second, political groupings with a unique moral vision.
Around town in Washington, DC, there’s a lot of chatter about better empowering the Department of Energy and its various investment arms. Some folks want it to have more flexible financing and investment authorities. Others want it to be more insulated from direct political influence. Others want to combine its capacities with those of other agencies. I do not think that these people realize that what they are asking for, in aggregate, is a public company. They want a company! An instrumentality with a technostructure! It’s that simple!
One of the greatest political economy papers of all time. Kalecki tha goat. 🐐🐐🐐
Here we should revisit footnote 11. When the New Industrial State was published in 1967, the United States was beginning its great descent into the stagflation that destroyed political support for American Keynesianism. The interaction between Galbraith’s and Kalecki’s theories looks quite interesting in this light. The ability of the technostructure to tolerate shifts in the costs of inputs and capital goods clearly did not translate to consumer price stability. The latter, as we know now, can de-range the former in a Kaleckian manner—not just by clogging up channels for translating consumer demand into demand for capital goods, but through what we might as well call “the political aspects of stagflation,” spoofing Kalecki himself: The average person in a shrinking industrial economy would rather feel security over their consumption possibilities than support a political economy that maintains the health of the industrial capital stock. A capital stock that can weather uncertainty might well be the price of crisis. (To be sure, the Volcker Shock was not a particularly healthy solution to stagflation, either.)
Janos Kornai is not one of Galbraith’s peers, but my colleague has previously reviewed his work and its relationship to Schumpeter, Kalecki, and Robinson. Implicitly, it touches on this discussion of Galbraith, too:
Under extended periods of low demand, it makes more sense to cut costs than to incur new ones by increasing productivity. Since low demand also means a loose labor market, it is simply cheaper to hire more people at a lower wage than innovate. Because, as Kornai acknowledges, prices are sticky, this can last for a very long time. In what some might call a "liquidity trap" the surplus economy can break down and instead of managing surplus, capitalist firms fight over a smaller pie. In doing so, they do not innovate, but rather live off rents and cheap labor. This is good for the individual capitalists who like the rents but it is not good for capitalism as a system. The innovative power of capitalism breaks down and so does Kornai’s typology [of capitalism as a “surplus economy”].
Please?
His actual politics seem to be, when read with today’s eyes, a fascinating and funny combination of abundance and degrowth.
At least two B2 bombers have flown over my house in DC as I have prepared this post for publication.
As per footnote 4.