Judge Posner (Reference Posner2010) treats organizational economics as a synonym for the principal-agent problem, as if the only issue arising in the design and evaluation of organizations was aligning the incentives of subordinates with the interests of their superiors. His perceptive discussion of a variety of private and public institutions through the lens of principal and agent illustrates not just the analytical power of this approach, but its limitations as well. Posner never defines the term organization explicitly, which leads him to see organizations where there are none, and in forcing the principal-agent template on social formations that are not organizations, he neglects other important aspects of their operation and the insights that might be gained from alternative conceptions of organizational economics.

More than 70 years ago, Ronald Coase and Friedrich Hayek, each in his own way, illuminated the contrasts between two kinds of social order, distinguished by the role in each of human purpose. Spontaneous orders, as Hayek called them, are created by the voluntary interaction of individuals in pursuit of their own purposes, constrained by rules of general application. Their hallmark is the sharp separation of individual purpose from systemic outcomes, epitomized by the process of efficient allocation that results, unintended by any buyer or seller, from their interaction in markets governed by the rule that every reallocation must be undertaken consensually. Especially when they are broadly beneficial to their participants, the intricate patterns of behavior and social institutions that emerge from these structured interactions often create the illusion of design, though in fact they are the products of no one's purpose or plan. To the extent that competitive markets perform the stupendous feats of information processing and behavioral coordination described by Hayek, the price theory that models them is itself an economics of organization, a theory of how people spontaneously organize themselves to allocate resources efficiently to individual preferences.

Individual purposes exist in centrally planned social orders as well, but they are subordinated to the achievement of one or a very few governing objectives. Unlike spontaneous orders, where a central plan is successful, the order that emerges from it is what it is because the planner intended that it be so and took the steps necessary to ensure that the individual subjects of the plan did what they had to do to bring this about. There need be nothing sinister about this imposition of governing purpose. Central plans may be administered consensually and received enthusiastically; the most successful ones almost always are. This is how we ordinarily think of organizations, as voluntary associations of individuals characterized by a governing objective and some sort of central plan designed to achieve it. Hayek (Reference Hayek1937: 52) believed that complex, unplanned order was the ‘central question of all social sciences’ and treated organizations accordingly, as useful foils for the study of spontaneous orders. But Coase opened a path to a robust economics of central planning, an economics of the organization that takes account of both the governing objective of the whole and the disparate purposes of the individual parts and has something to say about hierarchy and authority. Historians usually present the calculation debate of the interwar period as having pitted Mises and Hayek against Lange and others over whether a central plan to allocate efficiently to individual preferences could succeed in the absence of competitive prices, and count the economic collapse of the Soviet bloc as evidence for the Austrians. Yet in the midst of the controversy, Coase (Reference Coase1937: 388) pointed out that successful central planning is all around us, in the form of profit-seeking firms that replace spot markets in inputs with an ‘entrepreneur–coordinator’ who, within consensual limits, directs production by command.

An economics of organizations is thus an economics of central planning. Every planner, individual or central, must overcome a series of obstacles involving purpose, information, and control. First, an objective for the plan must be established, an outcome or set of criteria against which the plan's success can be tested. Then, enough information to construct a useful theory of the world relevant to the objective and apply it to the particular conditions of the moment must be acquired and used to devise a strategy to reach the goal. Finally, effective control over all the actors encompassed by the strategy must be achieved, to ensure that the planner's directives are carried out. These three parts of the planner's problem are generally interrelated, and their depth and complexity vary considerably from case to case. The nature and quantity of the information needed to formulate a strategy, and the planner's ability to rely upon or control the behavior of the people whose cooperation is essential to the plan's success, will often depend on the objective at stake, and vice versa, even where the plan is executed by a single person in her own interests. The crucial differences between individual planning in markets and central planning in organizations lie in the scope of the problems involved and the techniques available for solving them. In individual planning, objectives are personally defined and obstacles to them correspondingly local in character. The planner need only consult her own, subjective preferences and reactions to formulate a personal objective and measure the success of her strategy, and she need only inform herself of whatever she thinks she must know, in whatever form she finds useful, to execute it, but she can generally exercise control over others only through market exchange. In central planning, the planner must settle on a governing purpose that is both operationally feasible in light of the relevant information he can acquire and sufficiently straightforward to allow clear directives to be communicated, but he has, within continuously negotiable limits conditioned by his subjects’ trust in him and commitment to the governing objective, the power to control behavior by command.

As Posner suggests, the success of central planning turns largely on the willingness of subjects to subordinate their personal interests to those of the central planner or to see those interests as aligned with the planner's purpose, so they will be inclined to share what information they have with the planner and submit to his direction. This willingness can be achieved in several ways. One is by imbuing the subjects with a compelling sense of common purpose; another is through charismatic leadership that induces subjects to identify their own interests with those of the leader. Insofar as these approaches to the planner's problem seek its solution in persuading subjects to adopt a particular point of view or frame of mind, to want what they would otherwise not want, we may think of them as political rather than economic. But if subjects can pursue a purpose in common, so that each one sees realization of the planner's objective not as an end in itself but as a means to achieving her own personal ends, they can cooperate with the plan without surrendering the autonomy of their interests or identifying the planner's purpose as their own. Central planning organized in this way turns the political aspect of the earlier approaches into an economic one, a matter of coordinating the pursuit of many individual purposes rather than motivating the pursuit of a single, collective one. Because greater profit for the firm makes possible greater income for all its participants, it is just such a purpose in common, and, as Coase suggests, the ubiquity of the modern, profit-seeking business enterprise is testimony to the efficacy of consensual central planning in pursuit of a purpose in common (Adelstein, Reference Adelstein2005: 62–67).

From this perspective, it is not surprising that the largest part of Posner's essay is devoted to the alignment of incentives in large, profit-seeking corporations. These clearly are purposive organizations, and ones to which his principal-agent brand of organizational economics is especially well suited. In these firms, the purposes of a complex social system are reduced to maximizing a single, easily visualized number, in an environment where most of the things that determine profit, prices, costs, marginal products and the like, are themselves easily visualized (if often hard to ascertain) numbers. As a result, all the many tradeoffs that Posner shows to be involved in aligning incentives in the corporation are posed especially sharply. In theory, we know precisely what the firm's central planner needs to know to maximize profit, and can trace the interaction of most of the relevant variables, even as we ponder how to design organizations that put theory into practice. Indeed, where the relative costs and benefits of different organizational structures in various circumstances can be measured and compared, Posner's organizational economics becomes usefully normative, enabling him to tell corporations how, in light of their own situation, they can construct a central planning mechanism that in fact maximizes profits.

But as Posner makes clear, things get considerably murkier when the focus shifts from large corporations, with their sharp, quantifiable objectives and ‘high-powered’ pecuniary incentives, to government agencies, where objectives are less clearly defined, success is correspondingly more difficult to measure, and incentives for performance are often more moral than material. The FBI as Posner portrays it is a case in point. It too is a centrally planned organization, but one with two rather different objectives: the investigation of federal crimes, and domestic intelligence and the conduct of counterespionage and counterterrorism operations. The first of these goals offers quantifiable metrics of organizational success, arrest and conviction rates, the severity of sentences, the value of property recovered, and the like, which means that agents can generally be motivated by rewards based on objectively measurable individual performance. But the second does not – intelligence operations succeed when nothing rather than something happens, which makes success, and the specific contributions of individuals to it, significantly harder to assess. More importantly, as Posner shows, these two objectives seem to call for very different organizational structures, a geographically dispersed field-office system with looser central control for criminal work, and a more centralized national hierarchy for intelligence operations − and Posner has much of value to say about the incentive-alignment problems that this melding of objectives creates.

But as yet another approach to organizational economics, the capabilities perspective (cf. Langlois and Foss, Reference Langlois and Foss1999), suggests, the Bureau's structural difficulties extend beyond principals and agents. In this view, organizations are identified with the particular capabilities or competences they can achieve through the performance of evolved, organization-specific behavioral routines, with the size and shape of successful organizations determined more by the matching of individual and collective capabilities to the tasks at hand than by the costs of aligning incentives. How the organization solves the three-part planner's problem, that is, how it creates an internal order in the absence of prices, is the first and most important capability of any organization; as the organization operates in pursuit of its objectives, it develops a range of further capabilities related to the particular tasks it must perform to reach those goals. Organizations expand as existing capabilities are effectively applied to new operations in a way that serves the organization's objectives, and contract as capabilities once unique to the organization become cheaper to acquire through dealings with others in markets. Because the FBI's binary objectives have induced the evolution of two quite distinct sets of organizational capabilities, there is little room for the economies of scope that are typically realized when differing but complementary organizational objectives can be addressed by similar capabilities and much potential for organizational tension and dysfunction. In such a case, Posner and the capabilities theorists might agree, the Bureau would do well to spin off one or the other of its operations and concentrate its efforts on the task that remains.

Posner's discussion of the American ‘intelligence community’ is less satisfactory, for it is clear that this is not an organization, an association of individuals (or here, agencies) characterized by a single governing objective and a central plan to realize it, but an industry, an assortment of many organizations, each with a different governing objective and a different array of capabilities, operating at different points in the supply chain that produces intelligence for senior policymakers, to whom Posner assigns the role of principals with an especially unruly collection of agents. But the crucial problem here is not the alignment of incentives but the effective coordination of operations, the structuring of relationships between the various agencies such that each complements the activities of the others and the industry as a whole produces the necessary output with minimal friction. For this, the capabilities approach is a potentially rich source of insight. Posner notes that some of the intelligence agencies serve as suppliers of goods or services to others, and that the interests of one agency frequently conflict with those of another, a circumstance that makes the alignment of incentives across the industry hard to achieve and casts substantial doubt on the 1984 governmental reform that sought to unify the industry under the command of a Director of National Intelligence. A more effective approach to administering the industry would move incentives off center stage and focus instead on finding operational complementarities between the agencies and grouping them on the basis of compatible capabilities and objectives. Thus, the FBI's domestic intelligence division might achieve more were it more closely associated with the clandestine activities of the National Security Agency, the CIA, and similar espionage enterprises, not simply because they would no longer compete for limited budgetary resources to achieve the same general objectives but because this grouping would enable all the agencies to more easily pool (and improve) their complementary skills and realize economies of scope. The point is not that Posner's principal-agent analysis is incorrect or misleading – it is neither. But it is incomplete, in that a more catholic approach to organizational economics can open questions and answers inaccessible to an analyst concerned only with the alignment of incentives within unitary organizations.

Posner's final example, the American federal judiciary, is in many ways the most interesting of all. In some respects, those considered by Posner, it is an organization, albeit an unusual one. Judges all work for the same employer; there is a hierarchy of sorts in the judicial branch, though not one in which superiors closely supervise the work of underlings; and judges must be willing to do the work they are assigned in a way that satisfies their employer, so the question of incentive alignment is not a trivial one. But the organization's governing objective is not easy to pin down. Is it rendering the legally and factually correct decision in every case, and, if so, how could success be determined apart from the results of the judicial proceeding itself? Or is it consistently providing the public with the appearance of justice being done through fair procedures, whatever the outcome of particular cases might be? And just who is the principal to whom the judges are agents? In the case of the intelligence agencies, senior policymakers, and ultimately the President, are the obvious candidates, but neither the President nor Congress supervises the courts. Is it the Chief Justice, who administers the organization that employs the judges? Is it ‘the people,’ and, if so, why are they not given the power to elect or remove judges directly? Or is it the law itself, and, if so, how are we to determine when the law's purposes are or are not being well served? And how, precisely, is the behavior of judges related to the purposes of the organization and the interests of the principal, whomever that might be?

In other respects, however, the judiciary is not an organization at all. Posner explicitly excludes the Supreme Court from his analysis, but both the district and appellate courts have a substantial role to play in the development of the federal constitutional law, an exercise much more like the evolution of common law than the interpretation of statutes, particularly in cases where, as Posner argues rather implausibly in the case of European legislation, statutes essentially interpret themselves. To the extent that the purpose of the judiciary is to elaborate a detailed constitutional common law, it is achieved not within the confines of an organization but through a process that bears much closer resemblance to the operation of a spontaneous order, in which decentralized judges, influenced by local conditions and guided by precedent, interpret the vague language of the Constitution in particular factual situations, and appellate courts attempt to identify errors of interpretation and resolve important doctrinal disputes among the lower courts. No one knows what the perfect constitutional law would provide, so it is impossible to say whether the judicial process produces doctrine that approaches this ideal, and to perform their function properly judges must be given the freedom to decide cases as they see fit, with minimal direct supervision from above. The planner's problem is posed very differently when the central plan has a specific outcome in sight than when it seeks to nurture an order whose outcomes cannot be predicted. The designer of such a system is more like a gardener, creating conditions under which a decentralized, largely spontaneously ordered process can freely produce outcomes that no one can foresee, than a carpenter, a planner who employs a strategy or blueprint to realize a preconceived objective within the confines of a formal organization. For such a gardener, the organizational economics of greatest value is not Posner's, but Hayek's.