Review: People’s Republic of Walmart
People’s Republic of Walmart, by Leigh Phillips and Michal Rozworski, is the latest short text on a political theme to issue from Verso and Jacobin Magazine’s joint publishing venture. Its theme is planning, as it exists within capitalist firms today and might exist in the socialist polity of tomorrow.
The supercession of the market by economic planning was long thought, by both proponents and detractors, to be the characteristic goal of workers’ movements worldwide. Since the crisis of 2008, interest in the creeds of these movements (“socialism”, “communism”, “anarchism”) has exploded, due in no small part to the efforts of magazines like Jacobin and n+1.
The revival of these creeds has not been accompanied by a revival of interest in planning. This is a shame. Socialism without planning is Hamlet without the prince. As an effort to revive interest in economic planning generally, and the socialist calculation debate specifically, People’s Republic of Wal-Mart is laudable.
The authors take the reader on a tour of topics loosely related to the calculation debate per se and along the way present highly condensed arguments against capitalism and for economic planning with ‘democratic’ features1. What precisely is understood by ‘planning’ and what might make future implementations ‘democratic’ remains hazy throughout. Important arguments against planning are imperfectly communicated. As a result, People’s Republic is unlikely to persuade the skeptical reader of any political hue. This is lamentable, as the arguments in favor of certain conceptions of planning are much stronger than is commonly supposed today2.
Summary
According to Phillips and Rozworski, capitalism’s “irremediable, critical flaw” is that “the set of all useful things and the set of all profitable [to produce] things are not in perfect correspondence”3. New antibiotics could be developed and marketed to replace those that microbes have evolved to resist, which would save lives. They are not produced because it is not profitable to do so4. On the flip side, the production of fossil fuels is profitable despite “[undermining] human flourishing” by way of the greenhouse effect5.
This non-correspondence of useful and profitable-to-produce goods and services occurs because under capitalism “the primary method used to allocate things is the free market”6. Returning to the problem of antibiotic resistance to illustrate their point, the authors ask us to compare the incentives to produce a drug that merely amelioriates a chronic health condition with the incentives to produce a drug that cures the patient’s illness in one fell swoop. While the patient can be expected to prefer a cure, a revenue stream that endures as long as the patient does might prove more attractive to a drugmaker7. This anxiety about proper standards of conduct for the practioners of some art being subordinated to commercial standards is at least as old as the Platonic dialogues (“Is medicine to be the art of fee-making because the healer collects a fee for healing?”8).
The “allocation mechanism” of secondary importance in capitalist modernity is “planning”, which firms use to manage resources internally. Some firms, notably Walmart and and Amazon, are so “vast” that they are in every important respect “planned economies” like the former Soviet Union. “[When] we ask whether another [non-capitalist] world is possible, we are also asking: is there an alternate [primary] mechanism [with which] to allocate things?”9. The secondary mechanism must become primary, the “planning” used within firms today must be used without tomorrow10.
The former Soviet Union and Walmart - the “twin giants of central planning” - demonstrate that “planning works”, but they are normatively unacceptable (because undemocratic) models. We’re repeatedly assured that the kind planned economy envisioned by the authors would be “coordinated by ordinary working people, rather than by bureaucrats or bosses”.
We’re invited to imagine “…Walmart’s operational efficiency, its logistical genius, its architecture of agile economic planning… captured and transformed by those who aim toward a more egalitarian, liberatory society!”11 To create a viable socialist order, our fellow worker Prometheus must first steal fire from capital’s Mount Olympus.
Rebuttals to Austrian critiques
Having sketched the capitalist present, there are two arguments against the feasibility of socialism that the authors of People’s Republic are keen to rebut.
The first argument, attributed to Mises12, is that “planning at scale wasn’t feasible” because of the sheer number of calculations required to work out an allocative plan for a ‘large’ economy. They counter that the feasibility of planning at scale is established by the existence of capitalist enterprises like Walmart and Amazon that are “so vast that we should really call them centrally planned economies”. Like all capitalist enterprises, these buy and sell commodities (including labor services) in the market, but regulate their internal affairs through “planning” (rather than setting up internal markets for intermediate goods). Had anti-socialists (“Mises and friends”) been right, Walmart and Amazon would not exist because they would “long since have hit their wall of too many calculations to make”.
The second argument, attributed to Hayek, is that even if the calculations required to decide on an allocative plan were tractable, planners would not be able to collect “the data that goes into the equations”13. Phillips and Rozworksi counter that whatever the case may have been when Hayek wrote, the advent of “big data” in e-commerce, which “is testing the limits of how much granular information can be collected”, has eliminated any informational obstacles. Product recommendation systems like Amazon’s are held up as a “crucial innovation” for those who hope for a planned economy that can “satisfy consumer wants”, because they “[condense]… the chaos of individual tastes an opinions… into something useable”.
The inadequacy of these rebuttals
These rebuttals fly wide of their marks.
Walmart and Amazon have not overcome a once insuperable barrier to centrally planned allocation - computational or otherwise - because Walmart and Amazon do not allocate production goods.
Amazon and Walmart’s “logistical genius” - which is genuinely novel and interesting - is in solving the problem of distributing goods to consumers through a network of warehouses connected by variety transportation mechanisms. This problem is at root a series of enmeshed inventory, routing, and scheduling problems.
The animating question of the socialist calculation debate, on the other hand, was how a polity that had abolished private property in production goods (compare “capital”, “the means of production”, “higher order” goods) would allocate these production goods to different producing entities.
Solutions to the aforementioned logistical problems do not entail solutions to production good allocation problems or vice versa. Phillips and Rozworski often suggest otherwise (“manufacturing is merely another link in the retail supply chain…”). This is misleading insofar as it muddles the distinction between vertical integration - in which a firm begins producing inputs it once purchased, supplanting commodity exchange with a technical division of labor - and the complex subcontracting relationships typical of contemporary supply chains.
To see how today’s logistical problems and the economic planning problems of the calculation debate differ, consider how products you order on Amazon today are produced and sold.
Let’s say you order some “Amazon Basics”-branded #2 Pencils or a package of alkaline batteries. They are probably already resting in a warehouse near your home thanks to software that allows Amazon employees to anticipate the aggregate demand for these commodities. Because Amazon stores your purchase history and those of millions of other consumers, it possible for their employees to adjust the distribution of inventory across warehouses, the schedules of trucks, etc. They might even be able to indicate to some vendors that increased supply is likely to find demand through their platform. To do all of this at low cost is a vexing logistical problem, which Phillips and Rozworski are right to credit Amazon for tackling.
But how were our notional batteries and pencils manufactured?
Graphite powder, for example, is the raw material for both the anodes in alkaline batteries and the ‘lead’ in pencils. How did Amazon employees know how much graphite powder to allocate for the production of each? How were these allocations made consistent with the allocations of graphite needed to produce the myriad varieties of steel which constitute inputs for the production of hundreds of thousands of goods sold through Amazon?
Amazon employees didn’t know. They didn’t draw up ex ante consistent allocations of graphite. Amazon doesn’t allocate production goods to firms which manufacture for its virtual shelves. Those firms (and the firms they buy from, and the firms those firms buy from, etc.) turn to the market to acquire the inputs they believe will prove instrumental in their effort to turn a profit.
If Amazon or Walmart were systematically acquiring the firms that sold goods through their virtual storefront, replacing market exchange with a deliberate allocative plan and technical division of labor as they moved up the supply chain, there would be something to the book’s claim that “we should really call them centrally planned economies” of increasing and potentially “vast” scale. They are not, so there isn’t.
The Soviet Union’s state planning agency, Gosplan, did tackle problems of this kind through the method of “material balances” among others. Likewise, capitalist firms pursued vertical integration strategies to such extremes they too began to face similar administrative and computational difficulties. Henry Ford, for example, could boast that raw materials like coal and iron ore were transformed into a working automobile eighty-one hours later entirely within the confines of his River Rouge factory complex, with intermediate goods like sheet metal and cast metal parts produced along the way14. But there is no indication that these retail giants hope to begin allocating production goods to those who produce for their shelves. Why would they?
Phillips and Rozworski’s response to Hayek is similarly confused.
Hayek argued15 that there would be three types of information a central planning board would require but find impossible to marshal. First, a very granular and constantly updated inventory of all production goods, their state of repair (e.g. remaining useful lifetime of a machine) and location. Second, mathematical functions (think: lookup tables) summarizing the quantities of each input (e.g. X kilograms of steel tubing, Y drill presses, Z rubber tires, etc.) required to produce any quantity of a given commodity (e.g. N bicycles). Finally, “complete lists of the different quantities of all commodities which would be bought at any possible combination of prices of the different commodities which might be available” (i.e. an aggregate demand function) would have to be settled upon in order to begin calculating how best to make use of the aforementioned productive resources.
Perhaps Amazon’s product recommendation system could be extended to produce the third of Hayek’s three types of information. But it is evident that Amazon has not even taken the first step toward building systems to produce information of the first and second kinds, speculation about the perpetually arriving “internet of things” notwithstanding.
In sum, the inventory, scheduling, and routing techniques of e-commerce giants do not offer solutions to the problems Austrians identified with the planning proposals of “market socialism” that Rozworski and Phillips implicitly defend.
Planning for what?
Phillips and Rozworski’s enthusiasm for product recommendation systems suggests a vision of planning which owes more to the Austrian school than to Marxian socialism.
On their telling, Hayek saw markets as uniquely able to “aggregate and put to use the information dispersed throughout society”. But today, product recommendation systems have beat the market at its own game, so to speak (emphasis mine)16:
…Today’s deliberately planned IT systems are starting to create “group knowledge” (collective intelligence, or shared information that only emerges out of the interactions within or between groups of people) out of our individual needs and desires. And Amazon doesn’t just track market transactions. Beyond what you buy, the company collects data on what you browse, the paths you take between items, how long you stay on the page of each item you browse, what you place in your cart only to remove it later, and more.
Hayek could not have envisioned the vast amounts of data that can today be stored and manipulated outside of market interactions (and, to be fair, even many Marxists have assumed that the myriad capricious variables associated with faddish consumer items in particular forecloses the capacity for their socialization), although he certainly would have admired the capitalists such as Bezos who own the data and use it to pad their obscene fortunes. It is a delicious irony that big data, the producer and discoverer of so much new knowledge, could one day facilitate what Hayek thought only markets are capable of.
This “group knowledge” sounds suspiciously like the mysterious “catallaxy” appealed to by Austrians in their defenses of capitalism. In their eagerness to demonstrate that “big data” could possess the virtues Austrians ascribed to the market and the market alone, Phillips and Rozworski seem to have forgotten to question whether those were legitimate virtues in the first place.
For Hayek, capitalist society does not impose a collective purpose on its denizens, and is therefore a “Great Society” unprecedented in world history. The principal merit of this order is that it allows participants to coexist without having to discuss their aims, much less agree on them17:
The Great Society arose through the discovery that men can live together in peace and mutually benefit each other without agreeing on the particular aims that they severally pursue. The discovery that substituting abstract rules for obligatory concrete ends made it possible to extend the order of peace beyond the small groups pursuing the same ends, because it enabled each individual to gain from the skill and knowledge of others whom he need not even know and whose aims could be different from his own…
In the Great Society we all in fact contribute not only to the satisfaction of needs of which we do not know, but sometimes even to the achievement of ends of which we would disapprove if we knew about them. We cannot help this because we do not know for what purposes the goods or services which we supply to others will be used by them. That we assist in the realization of other people’s aims without sharing them or even knowing them, and solely in order to achieve our own aims, is the source of strength of the Great Society.
In fact, capitalist society does impose a collective project on its members: the endless accumulation of capital. Whatever goal we pursue - creating a great work of art, caring for the sick, housing our neighbors, etc. - we must either find a way turn a profit as we do so or secure the support of those who do (which support is sure to fail in times of crisis). Many on the socialist left today believe, as Hayek did, that our society does not impose a collective project on its members18. The thought follows naturally enough that socialist institutions must scrupulously preserve the market’s value-neutrality. But we cannot preserve what does not exist.
As the philosopher John O’Neill notes19, Hayek’s argument depends for its force on the assertion that “beliefs about values do not answer to rational argument”. If rational argument could settle disagreements about the products that needed to be provisioned and the services that needed to be offered for citizens to lead a variety of good lives, socialism might be both feasible and desirable. But because ideals and values are inaccessible to rational argument, a value-neutral mechanism is needed to “aggregate whatever ideals people happen to have, without conversation or judgement on those ideals themselves” (O’Neill). The market order of private property rights and free contract is just such a non-dialogical mechanism: an increase in the price of oil induces oil production regardless of suppliers’ attitudes toward climate change.
When Phillips and Rozworski argue that “big data” can “facilitate what Hayek thought only markets are capable of”, what can they have in mind but the creation of new mechanisms to aggregate preferences in lieu of political debate?
I am sure that such mechanisms can be constructed, and equally sure that they would bring us no closer to realizing socialist ideals.
The authors are explicit about the dual purpose of the book. They wish to rekindle interest in the socialist calculation debate (Pages 14-15) and to participate in that debate themselves (Page 16).↩
John O’Neill’s “The Market: Ethics, Knowledge, and Politics” (1998) is an excellent summary of these arguments, but was written for a specialist audience.↩
Page 4↩
A less contentious example than antibiotic resistance, about which experts seem to disagree, might have better served the authors in making this point.↩
Page 233↩
Page 9↩
This is my best effort at reconstructing the argument on page six. It is notable that here the authors appeal to a determinate non-commercial standard of “usefulness” - health - where in general they are disinclined to do so.↩
Plato’s Republic, Book I↩
Page 8↩
“… if allocation does not proceed via the market, then it will occur via economic planning, also know as ‘direct allocation’ …” (Page 10)↩
They take this theme from an essay by cultural critic Fred Jameson.↩
Mises did not argue that the essential difficulty with socialist planning would be the sheer number of calculations it would require. Other disputants, like Robbins and Hayek, put forward such “equation-difficulty” arguments, but only alongside others they held to be dispositive. The equation-difficulty argument was first hinted at by Barone in “The Ministry of Production in the Collectivist State” (1908). In “Human Action” (p. 706-711), Mises downplayed the significance of such arguments because he denied the premise that the relevant systems of equations could be formulated. The phrase “equation-difficulty argument” is from Rothbard, “The End of Socialism and the Calculation Debate Revisited” (1999).↩
Page 60↩
Henry Ford, “Today and Tomorrow” (1926). Google Books Preview.↩
Hayek, “Collectivist Economic Planning” (1935). See page 207 (in the essay “The Current State of the Debate”).↩
Page 85.↩
Hayek, “Law, Legislation and Liberty” (1982)↩
For more on the “liberal mirage” of a capitalism as a society without a collective purpose, see Murray, “The Mismeasure of Wealth” (Page 47).↩
O’Neill, “The Market” (1998), page 19.↩