Deficiencies of practical rationality in organizations

Suppose we are willing to take seriously the idea that organizations possess a kind of intentionality — beliefs, goals, and purposive actions — and suppose that we believe that the microfoundations of these quasi-intentional states depend on the workings of individual purposive actors within specific sets of relations, incentives, and practices. How does the resulting form of “bureaucratic intelligence” compare with human thought and action?

There is a major set of differences between organizational “intelligence” and human intelligence that turn on the unity of human action compared to the fundamental disunity of organizational action. An individual human being gathers a set of beliefs about a situation, reflects on a range of possible actions, and chooses a line of action designed to bring about his/her goals. An organization is disjointed in each of these activities. The belief-setting part of an organization usually consists of multiple separate processes culminating in an amalgamated set of beliefs or representations. And this amalgamation often reflects deep differences in perspective and method across various sub-departments. (Consider inputs into an international crisis incorporating assessments from intelligence, military, and trade specialists.)

Second, individual intentionality possess a substantial degree of practical autonomy. The individual assesses and adopts the set of beliefs that seem best to him or her in current circumstances. The organization in its belief-acquisition is subject to conflicting interests, both internal and external, that bias the belief set in one direction or the other. (This is the central thrust of experts on science policy like Naomi Oreskes.) The organization is not autonomous in its belief formation processes.

Third, an individual’s actions have a reasonable level of consistency and coherence over time. The individual seeks to avoid being self-defeating by doing X and Y while knowing that X undercuts Y. An organization is entirely capable of pursuing a suite of actions which embody exactly this kind of inconsistency, precisely because the actions chosen are the result of multiple disagreeing sub-agencies and officers.

Fourth, we have some reason to expect a degree of stability in the goals and values that underlie actions by an individual. But organizations, exactly because their behavior is a joint product of sub-agents with conflicting plans and goals, are entirely capable of rapid change of goals and values. Deepening this instability is the fluctuating powers and interests of external stakeholders who apply pressure for different values and goals over time.

Finally, human thinkers are potentially epistemic thinkers — they are at least potentially capable of following disciplines of analysis, reasoning, and evidence in their practical engagement with the world. By contrast, because of the influence of interests, both internal and external, organizations are perpetually subject to the distortion of belief, intention, and implementation by actors who have an interest in the outcome of the project. And organizations have little ability to apply rational rational standards to their processes of belief, intention, and implementation formation. Organizational intentionality lacks overriding rational control.

Consider more briefly the topic of action. Human actors suffer various deficiencies of performance when it comes to purposive action, including weakness of the will and self deception. But organizations are altogether less capable of effectively mounting the steps needed to fully implement a plan or a complicated policy or action. This is because of the looseness of linkages that exist between executive and agent within an organization, the perennial possibility of principal-agent problems, and the potential interference with performance created by interested parties outside the organization.

This line of thought suggests that organizational lack “unity of apperception and intention”. There are multiple levels and zones of intention formation, and much of this plurality persists throughout real processes of organizational thinking. And this disunity affects both belief, intention and action. Organizations are not univocal at any point. Belief formation, intention formation, and action remain fragmented and multivocal.

These observations are somewhat parallel to the paradoxes of social choice and various voting systems governing a social choice function. Kenneth Arrow demonstrated it is impossible to design a voting system that guarantees consistency of choice by a group of individual consistency voters. The analogy here is the idea that there is no organizational design possible that guarantees a high degree of consistency and rationality in large organizational decision processes at any stage of quasi-intentionality, including belief acquisition, policy formulation, and policy implementation. 

Exercising government’s will

Since the beginning of the industrial age the topic of regulation of private activity for the public good has been essential for the health and safety of the public. The economics of externalities and public harms are too powerful to permit private actors to conduct their affairs purely according to the dictates of profit and private interest. The desolation of the River Irk described in Engels’ The Condition of the Working-Class in England in 1844 was powerful evidence of this dynamic in the nineteenth century, and need for the protection of health and safety in the food industry, the protection of air and water quality, and establishment of regulations ensuring safe operation of industrial, chemical, and nuclear plants became evident in the middle of the twentieth century. (Of course it goes without saying that our current administration no longer concedes this point.)

A fundamental problem for understanding the mechanics of government is the question of how the will and intentions of government (policies and regulatory regimes) are conveyed from the sites of decision-making to the behavior of the actors whom these policies are meant to influence.

The familiar principal-agent problem designates precisely this complex of issues. Applying a government policy or regulation requires a chain of behaviors by multiple agents within an extended network of governmental and non-governmental offices. It is all too evident that actors at various levels have interests and intentions that are important to their choices; and blind obedience to commands from above is not a common practice within any organization. Instead, actors within an office or bureau have some degree of freedom to act strategically with regard to their own preferences and interests. What, then, are the arrangements that the principal can put in place that makes conformance by the agent more complete?

Further, there are commonly a range of non-governmental entities and actors who are affected by governmental policies and regulations. They too have the ability to act strategically in consideration of their preferences and interests. And some of the actions that are available to non-governmental actors have the capacity to significantly influence the impact and form of various governmental policies and regulations. The corporations that own nuclear power plants, for example, have an ability to constrain and deflect the inspection schedules to which their properties are subject through influence on legislators, and the regulatory agency may be seriously hampered in its ability to apply existing safety regulations.

This is a problem of social ontology: what kind of thing is a governmental agency, how does it work internally, and through what kinds of mechanisms does it influence the world around it (firms, criminals, citizens, local government, …)?

Two related ideas about the nature of organizations are relevant in this context. The idea of organizations as “strategic action fields” that is developed by Fligstein and McAdam (A Theory of Fields) fits the situation of a governmental agency. And the earlier work by Michel Crozier and Erhard Friedberg offer a similar account of the strategic action that jointly determines the workings of an organization. Here is a representative passage from Crozier and Friedberg:

The reader should not misconstrue the significance of this theoretical bet. We have not sought to formulate a set of general laws concerning the substance, the properties and the stages of development of organizations and systems. We do not have the advantage of being able to furnish normative precepts like those offered by management specialists who always believe they can elaborate a model of “good organization” and present a guide to the means and measures necessary to realize it. We present of series of simple propositions on the problems raised by the existence of these complex but integrated ensembles that we call organizations, and on the means and instruments that people have invented to surmount these problems; that is to say, to assure and develop their cooperation in view of the common goals.” 

L’acteur et le système, p. 11

(Here are some earlier discussions of these theories; link, link, link.  And here is a related discussion of Mayer Zald’s treatment of organizations; link.)

Also relevant from the point of view of the ontology of government organization is the new theory of institutional logics. Patricia Thornton, William Ocasio, and Michael Lounsbury describe new theoretical developments within the general framework of new institutionalism in The Institutional Logics Perspective: A New Approach to Culture, Structure and Process. Here is how they define their understanding of “institutional logic”:

… as the socially constructed, historical patterns of cultural symbols and material practices, including assumptions, values, and beliefs, by which individuals and organizations provide meaning to their daily activity, organize time and space, and reproduce their lives and experiences. (2)

The institutional logics perspective is a metatheoretical framework for analyzing the interrelationships among institutions, individuals, and organizations in social systems. It aids researchers in questions of how individual and organizational actors are influenced by their situation in multiple social locations in an interinstitutional system, for example the institutional orders of the family, religion, state, market, professions, and corporations. Conceptualized as a theoretical model, each institutional order of the interinstitutional system distinguishes unique organizing principles, practices, and symbols that influence individual and organizational behavior. Institutional logics represent frames of reference that condition actors’ choices for sensemaking, the vocabulary they use to motivate action, and their sense of self and identity. The principles, practices, and symbols of each institutional order differentially shape how reasoning takes place and how rationality is perceived and experienced. (2)

Here is a discussion of institutional logics; link.

So what can we say about the ontology of policy implementation, compliance, and executive decisions? We can say that —

  • it proceeds through individual actors in particular circumstances guided by particular interests and preferences; 
  • implementation is likely to be imperfect in the best of circumstances and entirely ineffectual in other circumstances; 
  • implementation is affected by the strategic non-governmental actors and organizations it is designed to influence, leading to further distortion and incompleteness. 

We can also, more positively, identify specific mechanisms that governments and executives introduce to increase the effectiveness of implementation of their policies. These include —

  • internal audit and discipline functions, 
  • communications and training strategies designed at enhancing conformance by intermediate actors, 
  • periodic purges of non-conformant sub-officials and powerful non-governmental actors, 
  • and dozens of other strategies and mechanisms of conformance.

Most fundamentally we can say that any model of government that postulates frictionless application and implementation of policy is flawed at its core. Such a model overlooks an ontological fundamental about government and other organizations, large and small: that organizational action is never automatic, algorithmic, or exact; that it is always conveyed by intermediate actors who have their own understandings and preferences about policy; and that it works in an environment where powerful non-governmental actors are almost always in positions to blunt the effectiveness of “the will of government”.

 

This topic unavoidably introduces the idea of corruption into the discussion (link, link). Sometimes the contrarian behavior of internal actors derives from private benefits offered them by outsiders influenced by the actions of government. (Hotels in Moscow?) More generally, however, it raises the question of conflicts of commitment, mission, role obligations, and organizational ethics.

%d bloggers like this: