1. Economic Calculation and the Limits of Organization

CHAPTER1

---

Economic Calculation and the Limits of Organization^†

---

Economists have become increasingly frustrated with the textbook model of the firm. The “firm” of intermediate microeconomics is a production function, a mysterious “black box” whose insides are off-limits to respectable economic theory (relegated instead to the lesser disciplines of management, organization theory, industrial psychology, and the like). Though useful in certain contexts, the textbook model has proven unable to account for a variety of real-world business practices: vertical and lateral integration, geographic and product-line diversification, franchising, long-term commercial contracting, transfer pricing, research joint ventures, and many others. As an alternative to viewing the firm as a production function, economists are turning to a new body of literature that views the firm as an organization, itself worthy of economic analysis. This emerging literature is the best-developed part of what has come to be called the “new institutional economics.¹ The new perspective has deeply enhanced and enriched our understanding of firms and other organizations, such that we can no longer agree with Ronald Coase’s 1988 statement that “[w]hy firms exist, what determines the number of firms, what determines what firms do... are not questions of interest to most economists” (Coase, 1988, p. 5). The new theory is not without its critics; Richard Nelson (1991), for example, objects that the new institutional economics tends to downplay discretionary differences among firms. Still, the new institutional economics—in particular, agency theory and transaction cost economics—has been the subject of increasing attention in industrial organization, corporate finance, strategic management, and business history.²

This chapter highlights some distinctive Austrian contributions to the theory of the firm, contributions that have been largely neglected, both inside and outside the Austrian literature. In particular, I argue that Mises’s concept of economic calculation—the means by which entrepreneurs adjust the structure of production to accord with consumer wants—belongs at the forefront of Austrian research into the nature and design of organizations. There is a unique Austrian perspective on economic planning, a perspective developed over the course of the socialist calculation debate. As was recognized in the early Austrian reinterpretations of the calculation debate (Lavoie, 1985; Kirzner, 1988a), Mises’s conception of the problem faced by socialist planners is part and parcel of his understanding of how resources are allocated in a market system. Mises himself emphasized that planning is ubiquitous: “[E]very human action means planning. What those calling themselves planners advocate is not the substitution of planned action for letting things go. It is the substitution of the planner’s own plan for the plans of his fellow men” (Mises, 1947, p. 493). All organizations plan, and all organizations, public and private, perform economic calculation. In this sense, the calculation problem is much more general than has usually been realized.

With their unique perspective on markets and the difficulties of resource allocation under central planning, third- and fourth-generation Austrian economists have always implicitly understood the economics of organization. In this context, as Nicolai Juul Foss (1994a, p. 32) notes, “it is something of a doctrinal puzzle that the Austrians have never formulated a theory of the firm.” Foss points out that many elements of the modern theory of the firm—property rights, relationship-specific assets, asymmetric information, the principal–agent problem—appeared, at least in elementary form, in Austrian writings since the middle stages of the calculation debate. Indeed, Rothbard’s treatment of firm size in Man, Economy, and State (1962) was among the first discussions to adopt explicitly the framework proposed by Ronald Coase in 1937, a framework that underlies most contemporary theorizing about the firm. Mises’s discussion in Human Action (1949) of the role of the financial markets foreshadows Henry Manne’s seminal 1965 article on the market for corporate control along with the recent recognition of finance as an essential part of economics.

Besides anticipating parts of the modern literature, Mises and Rothbard also introduced significant innovations, though this has not yet been generally recognized. Their contributions, while not part of a fully articulated, explicit theory of the firm, deserve attention and development, especially by those working on such issues from within the Austrian School.³ These contributions are Rothbard’s application of the calculation problem to the limits of the firm, and Mises’s discussion of how the financial markets both limit managerial discretion and perform the ultimate resource allocation task in a market economy.

The Textbook Theory of the Firm

In neoclassical economic theory, the firm as such does not exist at all. The “firm” is a production function or production possibilities set, a means of transforming inputs into outputs. Given the available technology, a vector of input prices, and a demand schedule, the firm maximizes money profits subject to the constraint that its production plans must be technologically feasible. That is all there is to it. The firm is modeled as a single actor, facing a series of relatively uncomplicated decisions: what level of output to produce, how much of each factor to hire, and so on. These “decisions,” of course, are not really decisions at all; they are trivial mathematical calculations, implicit in the underlying data. In the long run, the firm may also choose an optimal size and output mix, but even these are determined by the characteristics of the production function (economies of scale, scope, and sequence). In short: the firm is a set of cost curves, and the “theory of the firm” is a calculus problem.

To be sure, these models are not advertised as realistic descriptions of actual business firms; their use is purely instrumental. As David Kreps (1990b, p. 233)—himself much less sanguine about the merits of the traditional model than most—puts it: if real-world firms do not maximize profits as the traditional theory holds, “that doesn’t mean that profit maximization isn’t a good positive model. Only the data can speak to that, and then only after we see the implications of profit maximization for observable behavior.” However, even granting instrumentalism its somewhat dubious merits,⁴ the production-function approach is unsatisfactory, because it isn’t useful for understanding a variety of economic phenomena. The black-box model is really a theory about a plant or production process, not about a firm. A single firm can own and operate multiple production processes. Similarly, two or more firms can contract to operate jointly a single production process (as in a research joint venture). If we want to understand the scale and scope of the firm as a legal entity, then, we must look beyond the textbook model.

Coase and Transaction Costs

Ronald Coase, in his celebrated 1937 paper on “The Nature of the Firm,” was the first to explain that the boundaries of the organization depend not only on the productive technology, but on the costs of transacting business. In the Coasian framework, as developed and expanded by Williamson (1975, 1985, 1996), Klein, Crawford, and Alchian (1978), and Grossman and Hart (1986), the decision to organize transactions within the firm as opposed to on the open market—the “make or buy decision”—depends on the relative costs of internal versus external exchange. The market mechanism entails certain costs: discovering the relevant prices, negotiating and enforcing contracts, and so on. Within the firm, the entrepreneur may be able to reduce these “transaction costs” by coordinating these activities himself. However, internal organization brings another kind of transaction cost, namely problems of information flows, incentives, monitoring, and performance evaluation. The boundary of the firm, then, is determined by the tradeoff, at the margin, between the relative transaction costs of external and internal exchange. In this sense, firm boundaries depend not only on technology, but on organizational considerations; that is, on the costs and benefits of contracting.

The relative costs of external and internal exchange depend on particular characteristics of transaction: the degree to which relationship-specific assets are involved, the amount of uncertainty about the future and about trading partners’ actions, the complexity of the trading arrangement, and the frequency with which the transaction occurs. Each matters in determining the preferred institutional arrangement (that is, internal versus external production), although the first—”asset specificity”—is held to be particularly important. Williamson (1985, p. 55) defines asset specificity as “durable investments that are undertaken in support of particular transactions, the opportunity cost of which investments are much lower in best alternative uses or by alternative users should the original transaction be prematurely terminated.” This could describe a variety of relationship-specific investments, including both specialized physical and human capital, along with intangibles such as R&D and firm-specific knowledge or capabilities.

Economic Calculation and the Limits to Firm Size

Unfortunately, the growing economics literature on the theory of the firm focuses mostly on the costs of market exchange, and much less on the costs of governing internal exchange. The new research has yet to produce a fully satisfactory explanation of the limits to firm size (Williamson, 1985, chap. 6). In Coase’s words, “Why does the entrepreneur not organize one less transaction or one more?” Or, more generally, “Why is not all production carried on in one big firm?” (Coase, 1937, pp. 393–94). The theory of the limits to the firm is perhaps the most difficult and least well developed part of the new economics of organization. Existing contractual explanations rely on problems of authority and responsibility (Arrow, 1974); incentive distortions caused by residual ownership rights (Grossman and Hart, 1986; Hart and Moore, 1990; Hart, 1995); and the costs of attempting to reproduce market governance features within the firm (Williamson, 1985, chap. 6). It is here that Austrian theory has an obvious contribution to make, by applying Mises’s theorem on the impossibility of economic calculation under socialism. Rothbard has shown how the need for monetary calculation in terms of actual prices not only explains the failures of central planning under socialism, but places an upper bound on firm size.

The Socialist Calculation Debate: A Brief Review

To understand Mises’s position in the calculation debate, one must realize that his argument is not exclusively, or even primarily, about socialism. It is about the role of prices for capital goods. Entrepreneurs make decisions about resource allocation based on their expectations about future prices, and the information contained in present prices. To make profits, they need information about all prices, not only the prices of consumer goods but the prices of factors of production. Without markets for capital goods, these goods can have no prices, and hence entrepreneurs cannot make judgments about the relative scarcities of these factors. In short, resources cannot be allocated efficiently. In any environment, then—socialist or not—where a factor of production has no market price, a potential user of that factor will be unable to make rational decisions about its use. Stated this way, Mises’s claim is simply that efficient resource allocation in a market economy requires well-functioning asset markets. Because scholars differ about what Mises “really meant,” however, it may be useful here to provide a brief review of the debate.

Before 1920, according to the standard account,⁵ socialist theorists paid little attention to how a socialist economy would work in practice, most heeding Marx’s admonition to avoid such “utopian” speculation. Then Mises, known at the time mainly as a monetary theorist, published the sensational article later translated as “Economic Calculation in the Socialist Commonwealth” (1920).⁶ Mises claimed that without private ownership of the means of production, there would be no market prices for capital goods, and therefore no way for decision-makers to evaluate the relative efficiency of various production techniques. Anticipating the later argument for “market socialism,” Mises argued that even if there were markets for consumer goods, a central planner could not “impute” meaningful prices to capital goods used to produce them. In short, without market-generated prices for both capital and consumer goods, even the most dedicated planner would find it “impossible” to allocate resources according to consumer wants.

Throughout the 1920s and early 1930s Mises’s argument became the focus of intense discussion within the German-language literature. Eventually it was agreed that Mises was correct at least to point out that a socialist society could not do without such things as money and prices, as some early socialists had suggested, and that there was no feasible way to set prices according, say, to quantities of labor time. Nonetheless, it was felt that Vilfredo Pareto and his follower Enrico Barone (1908) had shown that nothing was “theoretically” wrong with socialism, because the requisite number of demand and supply equations to make the system “determinate” would exist under either capitalism or socialism. If the planners could somehow get the necessary information on preferences and technology, they could in principle compute an equilibrium allocation of final goods.

The most important response to Mises, however, and the one almost universally accepted by economists, was what became known as “market socialism” or the “mathematical solution,” developed by Fred Taylor (1929), H. D. Dickinson (1933), Abba Lerner (1934), and Oskar Lange (1936–37). In a system of market socialism, capital goods are collective property, but individuals are free to own and exchange final goods and services. The system would work like this. First, the Central Planning Board chooses arbitrary prices for consumer and capital goods. At those prices, the managers of the various state-owned enterprises are instructed to produce up to the point where the marginal cost of each final good is equal to its price, and then to choose the input mix that minimizes the average cost of producing that quantity. Then, consumer goods prices are allowed to fluctuate, and the Central Planning Board adjusts the prices of capital goods as shortages and surpluses of the final goods develop. Resources would thus be allocated according to supply and demand, through a process of “trial-and-error” essentially the same as that practiced by the managers of capitalist firms. Lange’s contribution, it has generally been held, was to show that production under market socialism could be just as efficient as production under capitalism, since the socialist planners “would receive exactly the same information from a socialized economic system as did entrepreneurs under a market system” (Heilbroner, 1970, p. 88).⁷

Market socialism was seen as an answer not only to Mises’s calculation problem, but also to the issue of “practicality” raised by Hayek and Lionel Robbins. Hayek, in his contributions to Collectivist Economic Planning (Hayek, 1935), later expanded in “The Competitive Solution” (1940) and his well-known papers “Economics and Knowledge” (1937) and “The Use of Knowledge in Society” (1945), and Robbins, in his The Great Depression (1934), had changed the terms of the debate by focusing not on the problem of calculation, but on the problem of knowledge. For Hayek and Robbins, the failure of socialist organization is due to a mechanism design problem, in that planners cannot allocate resources efficiently because they cannot obtain complete information on consumer preferences and resource availability. Furthermore, even if the planners were somehow able to acquire these data, it would take years to compute the millions of prices used by a modern economy. The Lange–Lerner–Taylor approach claimed to solve this preference-revelation problem by trial-and-error, so no actual computations would be necessary.⁸

With the widespread acceptance of the theory of market socialism, there developed an “orthodox line” on the socialist calculation debate, neatly summarized in Abram Bergson’s well-known survey of “Socialist Economics” (1948) and in Joseph Schumpeter’s Capitalism, Socialism and Democracy (1942, pp. 172–86). According to this line, Mises first raised the problem of the possibility of economic calculation under socialism, only to be refuted by Pareto and Barone; Hayek and Robbins then “retreated” to the position that socialist planners could calculate in theory, but that in practice the information problem would make this too difficult; then the market socialists showed that trial and error would eliminate the need for complete information on the part of the planners. Therefore, the argument goes, economic theory per se can say nothing conclusive about the viability of central planning, and the choice between capitalism and socialism must be purely political.

Calculation versus Incentives

The orthodox line on socialist planning has been modified in recent years with the development of incentive and information theory. The differences between capitalism and socialism, it is now typically held, lie in the different incentive properties of the two systems. Centrally directed systems are thought to be subject to greater agency costs—managerial discretion, shirking, and so on—than market systems (see, for example Winiecki, 1990). After all, Lange himself warned that “the real danger of socialism is that of a bureaucratization of economic life” (Lange, 1936–37, p. 109; italics in original).

As has been pointed out elsewhere (Rothbard, 1991, pp. 51–52), however, the calculation debate was not primarily about agency or managerial incentives. The incentive problem had long been known⁹ (if not fully developed) and was expressed in the famous question: “Under socialism, who will take out the garbage?” That is, if everyone is compensated “according to his needs,” what will be the incentive to do the dirty and unpleasant tasks; or, for that matter, any tasks at all? The traditional socialist answer was that self-interest is a product of capitalism, and that socialism would bring about a change in human nature. In the worker’s paradise would emerge a “New Socialist Man,” eager to serve and motivated only by the needs of his fellows. These early theorists seem to have assumed, to borrow the expression used by Oliver Williamson (1991a, p. 18) in a critique of a more recent socialist proposal, “the abolition of opportunism by agencies of the state.” Experience has exposed the charming naiveté of such notions.

But Mises’s challenge to socialism is distinct from this well-known incentive problem.¹⁰ Assume for the moment that everyone is willing to work just as hard under central direction as under a market system. There still remains the problem of exactly what directives the Central Planning Board will issue. The Board will have to decide what goods and services should be produced, how much of each to produce, what intermediate goods are needed to produce each final good, and so on. In a complex, modern economy with multiple stages of production, resource allocation requires the existence of money prices for capital goods, prices that under capitalism arise from an ongoing process of competitive bidding by entrepreneurs for the factors of production. This process cannot be replicated by input-output analysis, computer simulations, or any other form of artificial market. Mises’s main point was that socialism fails because decision makers require meaningful prices for all of these factors to choose from the vast array of possible factor combinations. “Without recourse to calculating and comparing the benefits and costs of production using the structure of monetary prices determined at each moment on the market, the human mind is only capable of surveying, evaluating, and directing production processes whose scope is drastically reduced to the compass of the primitive household economy” (Salerno, 1990a, p. 52).

The distinction between calculation and incentives is important because the modern economics literature on organizational design—from transaction cost explanations of firm size, to public choice theories of bureaucracy, to recent work on market socialism and the “soft budget constraint” (Kornai, 1986)—focuses primarily on incentive problems (possibly encouraged by Lange’s famous warning about bureaucracy). Incentive theory asks how, within a specified relationship, a principal can get an agent to do what he wants him to do. Mises’s problem, however, was different: How does the principal know what to tell the agent to do? That is, just what activities ought to be undertaken? What investments should be made? Which product lines expanded and which ones contracted? The ideas developed in the calculation debate suggest that when organizations are large enough to conduct activities that are exclusively internal—so that no reference to the outside market is available—they will face a calculation problem as well as an incentive problem.

In this sense, market-socialist proposals are mostly irrelevant to the real problems of socialist organization. This is the case Mises himself sought to make in his critique of market socialism in Human Action (Mises, 1949, pp. 694–711). There he complained that the market socialists—and, for that matter, all general equilibrium theorists—misconceive the nature of “the economic problem.” Lange, Lerner, and Taylor looked primarily at the problem of consumer goods pricing, while the crucial problem facing a modern economy concerns the capital structure: namely, in what way should capital be allocated to various activities? The market economy, Mises argued, is driven not by “management”—the performance of specified tasks, within a framework given to the manager—but by entrepreneurship, the speculation, arbitrage, and other risk-bearing activities that determine just what the managerial tasks are. It is not managers but entrepreneurs, acting in the capital and money markets, who establish and dissolve corporations, create and destroy product lines, and so on. These are precisely the activities that even market socialism seeks to abolish. In other words, to the extent that incentives are important, what socialism cannot preserve are high-powered incentives not in management, but in entrepreneurial forecasting and decision making.

Mises has been described as saying that it is unreasonable to expect managers of socialist enterprises to “play market,” to act as if they were managers of private firms where their own direct interests were at stake. This may be true, but Mises’s prime concern was that entrepreneurs cannot be asked to “play speculation and investment” (Mises, 1949, p. 705). The relevant incentive problem, he maintains, is not that of the subordinate manager (the agent), who takes the problem to be solved as given, but that of the speculator and investor (the principal), who decides just what is the problem to be solved. Lange, Lerner, and Taylor see the market through a strictly static, neoclassical lens, where all the parameters of the system are given and only a computational problem needs to be solved. In fact the market economy is a dynamic, creative, evolving process, in which entrepreneurs—using economic calculation—make industries grow and shrink, cause new and different production methods to be tried and others withdrawn, and constantly change the range of available products. It is these features of market capitalism, and not the incentives of agents to work hard, that are lost without private property ownership.

Indeed, traditional command-style economies, such as that of the former USSR, appear to be able only to mimic those tasks that market economies have performed before; they are unable to set up and execute original tasks.

The [Soviet] system has been particularly effective when the central priorities involve catching up, for then the problems of knowing what to do, when and how to do it, and whether it was properly done, are solved by reference to a working model, by exploiting what Gerschenkron... called the “advantage of backwardness.” ...Accompanying these advantages are shortcomings, inherent in the nature of the system. When the system pursues a few priority objectives, regardless of sacrifices or losses in lower priority areas, those ultimately responsible cannot know whether the success was worth achieving. The central authorities lack the information and physical capability to monitor all important costs—in particular opportunity costs—yet they are the only ones, given the logic of the system, with a true interest in knowing such costs. (Ericson, 1991, p. 21).

Without economic calculation, there is no way to figure out if tasks have been performed efficiently. Hence without markets for physical and financial capital—which determine what tasks will be performed and whether they have been performed adequately—an economic system has difficulty generating anything new, and must rely on outside references to tell it what to do. Of course, the only reason the Soviet Union and the communist nations of Eastern Europe could exist at all is that they never fully succeeded in establishing socialism worldwide, so they could use world market prices to establish implicit prices for the goods they bought and sold internally (Rothbard, 1991, pp. 73–74). In Mises’s words, these economies

were not isolated social systems. They were operating in an environment in which the price system still worked. They could resort to economic calculation on the ground of the prices established abroad. Without the aid of these prices their actions would have been aimless and planless. Only because they were able to refer to these foreign prices were they able to calculate, to keep books, and to prepare their much talked about plans. (Mises, 1949, pp. 698–99).

As we will see below, the firm is in the same situation: it needs outside market prices to plan and evaluate its actions.

Rothbard and the Limits of Organization

Rothbard’s main contribution to the theory of the firm was to generalize Mises’s analysis of the problem of resource allocation under socialism to the context of vertical integration and the size of the organization. Rothbard writes in Man, Economy, and State that up to a point, the size of the firm is determined by costs, as in the textbook model. But “the ultimate limits are set on the relative size of the firm by the necessity for markets to exist in every factor, in order to make it possible for the firm to calculate its profits and losses” (Rothbard, 1962, p. 599). This argument hinges on the notion of “implicit costs.” The market value of opportunity costs for factor services—what Rothbard calls “estimates of implicit incomes”—can be determined only if there are external markets for those factors (pp. 607–09). For example, if an entrepreneur hires himself to manage his business, the opportunity cost of his labor must be included in the firm’s costs. But without an actual market for the entrepreneur’s managerial services, he will be unable to figure out his opportunity cost; his balance sheets will therefore be less accurate than they would if he could measure his opportunity cost.

The same problem affects a firm owning multiple stages of production. A large, integrated firm is typically organized as groups of semiautonomous business units or “profit centers,” each unit or division specializing in a particular final or intermediate product. The central management of the firm uses the implicit incomes of the business units, as reflected in statements of divisional profit and loss, to allocate physical and financial capital across the divisions. More profitable divisions are expanded, while less profitable divisions are scaled back. Suppose the firm has an upstream division selling an intermediate component to a downstream division. To compute the divisional profits and losses, the firm needs an economically meaningful “transfer price” for the component. If there is an external market for the component, the firm can use that market price as the transfer price.¹¹ Without a market price, however, a transfer price must be estimated in another way.

In practice, this is typically done on a cost-plus basis; sometimes, the buying and selling divisions are left free to bargain over the price (Eccles and White, 1988; Shelanski, 1993; King, 1994). At the very least, any artificial or substitute transfer prices will contain less information than actual market prices; Rothbard (1962, p. 613) puts it more strongly, calling a substitute price “only an arbitrary symbol.” In either case, firms relying on these prices will suffer. “Not being able to calculate a price, the firm could not rationally allocate factors and resources from one stage [or division] to another” (Rothbard, 1962, p. 613) The use of internally traded intermediate goods for which no external market reference is available introduces distortions that reduce organizational efficiency. This gives us the element missing from contemporary theories of economic organization, an upper bound: the firm is constrained by the need for external markets for all internally traded goods. In other words, no firm can become so large that it is both the unique producer and user of an intermediate product; for then no market-based transfer prices will be available, and the firm will be unable to calculate divisional profit and loss and therefore unable to allocate resources correctly between divisions. As Rothbard puts it:

Since the free market always tends to establish the most efficient and profitable type of production (whether for type of good, method of production, allocation of factors, or size of firm), we must conclude that complete vertical integration for a capital-good product can never be established on the free market (above the primitive level). For every capital good, there must be a definite market in which firms buy and sell that good. It is obvious that this economic law sets a definite maximum to the relative size of any particular firm on the free market.... Economic calculation becomes ever more important as the market economy develops and progresses, as the stages and the complexities of type and variety of capital goods increase. Ever more important for the maintenance of an advanced economy, then, is the preservation of markets for all the capital and other producers’ goods. (Rothbard, 1962, p. 613; italics in original)

Like the centrally planned economy, the firm needs market signals to guide its actions; without them the firm cannot survive. Note that in general, Rothbard is making a claim only about the upper bound of the firm, not the incremental cost of expanding the firm’s activities (as long as external market references are available). As soon as the firm expands to the point where at least one external market has disappeared, however, the calculation problem exists. The difficulties become worse as more and more external markets disappear, as “islands of noncalculable chaos swell to the proportions of masses and continents. As the area of incalculability increases, the degrees of irrationality, misallocation, loss, impoverishment, etc., become greater” (p. 548). In other words, the firm is limited by the extent to which markets exist for the goods it allocates internally. Without market prices for these goods, the firm must rely on relatively costly and inefficient methods of generating its own accounting prices, to perform internal calculations.¹²

Significantly, it is at this point in the discussion in Man, Economy, and State (p. 548) that Rothbard launches into a discussion of the socialist calculation debate, making it obvious that the two issues are inextricably linked. The reason that a socialist economy cannot calculate is not that it is socialist, but because a single agent owns and directs all resources. Expanding on this point in his 1976 essay on “Ludwig von Mises and Economic Calculation Under Socialism,” Rothbard explains:

There is one vital but neglected area where the Mises analysis of economic calculation needs to be expanded. For in a profound sense, the theory is not about socialism at all! Instead, it applies to any situation where one group has acquired control of the means of production over a large area—or, in a strict sense, throughout the world. On this particular aspect of socialism, it doesn’t matter whether this unitary control has come about through the coercive expropriation brought about by socialism or by voluntary processes on the free market. For what the Mises theory focuses on is not simply the numerous inefficiencies of the political as compared to the profit-making market process, but the fact that a market for capital goods has disappeared. This means that, just as socialist central planning could not calculate economically, no One Big Firm could own or control the entire economy. The Mises analysis applies to any situation where a market for capital goods has disappeared in a complex industrial economy, whether because of socialism or because of a giant merger into One Big Firm or One Big Cartel. (Rothbard, 1976, p. 75)

The Mises analysis thus applies to any situation where the market for a particular capital good disappears because a firm has become so large that it is the unique producer and user of that capital good. As we have seen, such a firm will not be viable.

It is surprising that Rothbard’s extension of Mises’s argument has received virtually no attention in the Austrian literature, even though the point appears four times in Man, Economy, and State (p. 536, p. 543, pp. 547–48, and p. 585) and again in the 1976 essay.¹³ The argument needs further development and elaboration, which should prove a useful exercise because the contemporary literature on the size of the firm lacks an adequate explanation for the limits to organization. The Rothbard analysis also suggests a line of research in business strategy: all else equal, firms able to use market-based transfer prices should outperform, in the long run, firms using administered or negotiated transfer prices.¹⁴ As of yet, there is little empirical work on this topic, despite a growing interest in Austrian economics within the strategic management field (Jacobson, 1992; Lewin and Phelan, 1999; Foss and Mahnke, 2000; Langlois, 2001; Roberts and Eisenhardt, 2003; Yu, 2003; Ng, 2005; Mathews, 2006).

A related issue that has received considerable attention, however, is the difficulty of allocating overhead or fixed cost across divisions. If an input is essentially indivisible (or nonexcludable), then there is no way to compute the opportunity cost of just the portion of the input used by a particular division (see Rogerson, 1992, for a discussion of these problems).¹⁵ Firms with high overhead costs should thus be at a disadvantage relative to firms able to allocate costs more precisely between business units. Indeed, in the literature on cost accounting there has been some recent interest in “market simulation accounting” (Staubus, 1986), by which firms try to assess the price at which an asset would trade in an active market, based on observed market prices and related information. The Rothbardian position on the limits to firm size suggests that the market simulation approach may prove a useful accounting technique.

By the time of the 1976 paper, Rothbard had adopted an explicitly Coasian framework in his discussion of the limits to firm size. His own treatment, Rothbard says,

serves to extend the notable analysis of Professor Coase on the market determinants of the size of the firm, or the relative extent of corporate planning within the firm as against the use of exchange and the price mechanism. Coase pointed out that there are diminishing benefits and increasing costs to each of these two alternatives, resulting, as he put it, in an “ ‘optimum’ amount of planning” in the free market system. Our thesis adds that the costs of internal corporate planning become prohibitive as soon as markets for capital goods begin to disappear, so that the free-market optimum will always stop well short not only of One Big Firm throughout the world market but also of any disappearance of specific markets and hence of economic calculation in that product or resource. (Rothbard, 1976, p. 76)

This is noteworthy because even as late as 1972, Coase was describing his 1937 paper as “much cited and little used” (Coase, 1972, p. 62). Alchian and Demsetz’s “Production, Information Costs, and Economic Organization” came out only in 1972, and Williamson’s Markets and Hierarchies in 1975. Rothbard was thus among the earliest writers to develop and extend the Coasian perspective.

Alternative Austrian Approaches

There is some debate within the Austrian literature about whether the basic Coasian approach is compatible with Austrian economics. O’Driscoll and Rizzo (1985, p. 124), while acknowledging Coase’s approach as an “excellent static conceptualization of the problem,” argue that a more evolutionary framework is needed to understand how firms respond to change. Some Austrian economists have suggested that the Coasian framework may be too narrow, too squarely in the general-equilibrium tradition to deal adequately with Austrian concerns (Boudreaux and Holcombe, 1989; Langlois, 1994a). They contend that the contemporary theory of the firm, following Coase, retains the perspective of static equilibrium analysis and profit maximization over a fixed set of outcomes with known probabilities. As an alternative, some writers propose the framework in Frank Knight’s Risk, Uncertainty, and Profit (1921). The Knightian framework, they argue, offers genuine uncertainty, disequilibrium and process analysis, and thus a scope for real entrepreneurship—aspects purportedly more congenial to Austrians. “The Coasian and Knightian theories of the firm deal with the issue [of the existence of firms] from two different vantage points. The Coasian theory takes the inputs and outputs in the firm’s production process as given, and models the firm as an organization that acts to minimize the costs of transforming these inputs into outputs.... However, in Knight’s model, entrepreneurship is the primary role of the firm (Boudreaux and Holcombe, 1989, p. 152). Williamson’s transaction cost economics, as characterized by Langlois (1994a, p. 175), does broaden the notion of cost minimization to include transaction costs as well as production costs, but it remains essentially a static exercise with a limited role for expectations: “Seldom does the theory give thought to the possibility that organizational forms may be influenced as much by environments that exist only as future possibilities, imagined or feared.”

To be sure, the Knightian concept of the profit-seeking entrepreneur, investing resources under uncertainty, is one of the great contributions to the theory of the firm. As discussed in chapters 4, 5, and 6 below, it is close to Mises’s concept of the entrepreneur (closer, in my view, than Israel Kirzner’s understanding of entrepreneurship). Still, these critiques of the Coasian framework paint with too broad a brush; as Foss (1993c) points out, there are “two Coasian traditions.” One tradition, the nexus-of-contracts branch associated with Alchian and Demsetz (1972), studies the design of ex ante mechanisms to limit shirking when supervision is costly. Here the emphasis is on monitoring and incentives in an (exogenously determined) moral-hazard relationship. The aforementioned criticisms may apply to this branch of the modern literature, but they do not apply to the other tradition, the governance or asset-specificity branch, especially in Williamson’s more heterodox formulation. Williamson’s transaction cost framework incorporates non-maximizing behavior (bounded rationality); true, “structural” uncertainty or genuine surprise (complete contracts are held not to be feasible, meaning that all ex post contingencies cannot be contracted upon ex ante); and process or adaptation over time (trading relationships develop over time, typically undergoing a “fundamental transformation” that changes the terms of trade). In short, “at least some modern theories of the firm do not at all presuppose the ‘closed’ economic universe—with all relevant inputs and outputs being given, human action conceptualized as maximization, etc., that [some critics] claim are underneath the contemporary theory of the firm” (Foss, 1993a, p. 274). Stated differently, one can adopt an essentially Coasian perspective without abandoning the Knightian or Austrian view of the entrepreneur as an uncertainty-bearing, innovating decision-maker.¹⁶

Similarly, the approach described in this chapter differs from that advanced in the literature on “market-based management” (Ellig, 1993; Ellig and Gable, 1993; Koch, 2007). Market-based management is the philosophy that firm success depends critically on the ability to replicate marketlike features within the organization. One of these is “internal markets” for intermediate goods (and services such as financial, legal, accounting, and R&D support) along with the establishment of strict profit-center divisions. Like market prices, these internal prices convey information about local circumstances. Other features include an explicit “mission” or recognition of the firm’s core competence, clearly defined roles and responsibilities for lower-level employees (analogous to property rights in a market economy), employee rewards based on performance (a profit-and-loss system), a well-defined “corporate culture” (customs, behavioral norms), and decentralized decision making.

Underlying market-based management is the team-production or nexus-of-contracts model of the firm advanced by Alchian and Demsetz (1972), supplemented with the “capabilities” theory of Edith Penrose (1959), G. B. Richardson (1972), David Teece (1980; 1982), and others. But the market-based management literature, like other writings in the nexus-of-contracts tradition, appears to mischaracterize the nature of “planning” within the firm. For example, it attributes to the Coase-Williamson tradition the view that “internal markets are doomed to failure, because the business firm is by nature a command hierarchy” (Ellig, 1993, p. 9). The Coasian tradition, however, does not imply that firms do or should adopt a command-and-control structure; on the contrary, as we have already seen, the modern firm will tend to be significantly decentralized, so that managers and workers at all levels of operations can make use of local knowledge. All decisions are not made from above, by executive fiat; the “M-form” corporation described by Williamson and Chandler is a blend of market and hierarchy, of centralization and decentralization.

In other words, the entrepreneur does make some decisions by “fiat”; the firm is definitely a taxis, rather than a cosmos (to use Hayek’s esoteric terminology).¹⁷ This does not imply, however, that all decisions must be made from the top; we can agree with the market-based management literature that “neither central planning nor command-and-control are the defining characteristics of a business firm” (Ellig, 1993, p. 11). Indeed, given competition in the product and factor markets, firms will always tend to select the optimum amount of “market-like” features. The firm’s problem, then, is not too much “conscious” planning; the crucial issue is whether these plans are made, and tested, from within a larger market setting. The entrepreneur’s plans can be carried out, as we saw above, only when there are definite markets for all internally traded goods or activities. What firms need is not necessarily internal markets, but the information generated by market prices.

Conclusion

This chapter has highlighted some Austrian contributions to the theory of the firm and suggested directions for future research along the same lines. In particular, Rothbard’s argument about the need for markets in intermediate goods, and how that places limits on the scale and scope of the organization, deserves further development. The chapter also points the way toward an Austrian approach that makes the entrepreneur, and his acts of resource allocation using monetary calculation, central to the theory of the firm.