Hayekian Knowledge Arguments: An Epistemic Fallacy?
| Nicolai Foss |
A small handful of papers have become highly influential in economics as well as in management and organization research. One such paper is Friedrich Hayek’s 1945 essay, “The Use of Knowledge in Society,” a paper that emerged in the context of the debate on the viability and efficiency of planned resource allocation on the societal level (i.e., socialism) that raged among academic economists in (particularly) the inter-war period. Hayek famously argued that planning confronts inherent knowledge-based constraints, and these constraints are certainly binding at a scale of activity that makes comprehensive overall management/planning of economy-wide resource allocation deeply inefficient. Many modern management thinkers have echoed this argument, arguing that “traditional” authority relations are increasingly challenged by the (increasingly) dispersed nature of knowledge.
However, at least when applied to authority in firms the Hayekian knowledge argument arguably misconstrues the nature of managerial authority, because it is based on an epistemic fallacy.
The fallacy lies in the argument that making rational decisions as a principal or employer concerning agents or employees presupposes full knowledge of their available action repertoire. As Simon argues authority can be understood as a command that takes the form of a result to be produced, a principle to be applied, or goal constraints, so that while the end result may be described, the methods of reaching it are not. In actuality, this is no different from how most agents interact with specialized other agents in the market place. Thus, most of us non-plumbers usually don’t know the full action set of the plumber, or what exactly to do in response to some calamity involving the plumbing of our house. Still, we do rely on our judgment of the plumber, possibly supplemented with reputational mechanisms. Market transactions do not break down when there is only partial knowledge overlap. By the same token, hierarchical mechanisms, involving authority, do not break down when there is only partial knowledge overlap. Exchange, whether in the market or inside firms, is perfectly possible in the face of ignorance.
Frank Knight in Risk, Uncertainty and Profit very clearly recognized this. The arguably key concept in Knight is that of “judgment,” that is, the ability to make decisions concerning resource uses in situations involving genuine uncertainty, that is, situations in which the range of outcomes and the probabilities of outcomes are unknown. More compactly, judgment is the human faculty that makes it possible for us to make decisions, even under severe ignorance. While Knight’s (1921) main interest in judgment lies in its role as a basis for a theory of profit and the firm, he understood very clearly that the effective exercise of authority does not require full knowledge of an employee’s action set and precise knowledge of exactly which action should be picked in response to contingencies: “What we call ‘control’ consists mainly of selecting someone else to do the ‘controlling.’ Business judgment is chiefly judgment of men. We know things by knowledge of men who know them and control things in the same indirect way” (Knight 1921: 291). Delegation, Knight argues, rests on judgment. Again, such judgment does not require that that the principal knows the employee’s entire action set. Hierarchical organization is conceptualized by Knight as a layer of agents exercising judgment of subordinates’ judgment, that is, nested judgment. Although a manager may still be ignorant in an important sense about the knowledge controlled by his employees he does not suffer from complete ignorance; there is some, possibly very modest, knowledge overlap. He may therefore be able to pass judgment on the overall abilities of those employees, and, in particular, about how actions based on their knowledge may be coordinated. In other words, it is possible to have knowledge of types of interdependencies between actions based on different knowledge elements without possessing much knowledge of the actual interdependencies or the actions themselves. (Whether this has implications for the theory of socialism is an open issue).
The above issues are treated in a recent paper, “Hayekian Knowledge Arguments in Organizational Theory” by Kirsten Foss and yours truly. Mail me at email@example.com if you want to receive a copy.