2 - Reconsidering the Good of Improving Accountability

Summary

Among the rhetorical pleas that follow most instances of public dissatisfaction is the call for more or better accountability. Accountability is a lauded notion, a “golden concept” that is considered widely as critical to the success of democratic government. Such pleas, I will argue, are misplaced. Rather than starting from the premise of accountability as an idea that no one can be against, I consider the possibility that accountability undermines the very notion it ostensibly promotes: self-government. The concept of accountability in modern political theory is tied more closely to the emergence of an impersonal administrative state than it is to the hopeful horizon of a democratic one. In practice and in theory, it is a concept of irresponsibility, a technological approach to government that provides the comforts of impersonal rationality.

2.1 Introduction

One of the more curious developments of twentieth-century modern political thought is the laudatory treatment of accountability. Despite frequent debates over accountability, the general sense is that it is something to be preserved and improved for the sake of democracy.¹ Accountability has long been connected to democracy in the sense of representation, a feature that dates to JS Mill. But the modern notion of accountability is different in substance from the liberal idea, which is tied to representation. Accountability in its modern guise trends away from ideas of representation towards matters of technology and design. What I mean by this is that accountability is a derivative value that functions at the level of instrumentation. What makes accountability good is its use in achieving a goal. Accountability is still regarded as and spoken about as a political good, but this is generally a remnant of the liberal idea of accountability tied to representation. As Mill put it, perfecting accountability, in the sense that he meant it, depended on aligning the interests of rulers with that of the people.² Mill, however, did not have the insight into what accountability looks like in a modern bureaucratic state, which is the modern version upon which I will focus here.

The approbation of accountability derives from a sense that accountability is a promotional good of democracy. In other words, accountability has been framed over the past century as a feature of modern government that is essential to the broad success of realizing democratic government, which has lately been transformed into an idea of participatory governance. Achieving accountability entails success in both realizing the increasingly complex objectives of the modern state and satisfying the values of democracy. As such, many regard promoting accountability as de rigueur of democracy. The idea carries a normative force that appears essential to addressing imbalances of power that favor experts against the public. As Dawson and Maricut-Akbik describe in their introductory chapter to this volume, the normative goods of accountability can help guide frameworks to address the nongovernmental, extra-nation-state institutions developed and developing alongside the increasingly complex demands of modern governance. They propose four normative goods: openness, nonarbitrariness (procedural limits on discretion), effectiveness through measured performance, and ensuring that actions are in the public interest (publicness). These goods underscore both the procedural and substantive qualities of accountability. Procedural accountability ensures that activities are done correctly (open, public, nonarbitrary) and lead to appropriate outcomes (effective). Substantive accountability places higher demands on the institutional setting, as Dawson and Maricut-Akbik describe, by requiring an explanation of the decisions behind the activities.

It is unusual not to be swept along in the laudation of accountability. I myself have described accountability in two previous essays, one in which I attempt to deepen the connection between accountability and democracy by incorporating practices of contestation and nondomination into its conceptual fold.³ I see such matters slightly differently now, especially considering some of the distinctions I will address in this essay, namely the distinction between responsibility and accountability and the ways in which accountability serves bureaucracy and not democracy. I focus upon the concept as it applies to public administration and the political formations associated with it (i.e. the administrative state). The administrative idea of accountability is tied to design and performance, evident in the procedural sense of accountability, making accountability a concept that deepens technological systems of control. I also identify problems on the substantive side of accountability where actors are required to tell a story (give an account) of their decisions. The account itself is plagued by the dual problem that, first, it is the account itself that is subjected to procedural accountability through design and, second, that accountability as a practice does not permit a meaningful discourse between experts and the public. Ultimately, accountability is a practice that formalizes expertise into governing, an idea contrary to the prevailing notion of it as a value that ensures democratic control.

I doubt the integrity of ideas that connect modern accountability to democracy and see the rethinking of accountability against democracy as an important political problem of our time. Accountability is not an idea that promotes democracy. Contra democracy, the notion of accountability is found in strengthening bureaucracy, and it does so without explicit regard for ideas of democracy. Viewing accountability not as an institutional good but as a concept with relational consequences achieved through technology brings into question this largely uncriticized connection between democracy and accountability. In this essay, I question “the good” of accountability and identify how the concept of accountability in both the procedural and substantive sense reinforces the power of experts above the public.

2.2 Responsibility and Accountability

I begin by distinguishing between accountability and the more common notion of responsibility. Responsibility partly entails giving an account of one’s actions, meaning an explanation for what one did, why one did it, and what reasonable expectations were behind doing it. This does not mean that responsibility is the account itself. Responsibility is something that one can have in the sense of “my responsibility.” We can speak of responsibility in a way that gives it a sense of character and connection with an individual.

On the other hand, accountability takes that element of the account as everything. For accountability, the details of the account prevail, so much so that in its reductive sense, accountability renders pointless the thinking and decision that lies behind the human act itself. One cannot speak of “my accountability.” It is meaningless. Being accountable means being held accountable. This contrasts sharply with the quality of responsibility, which does not require an other to act upon me. One does not need another to make one responsible. The other can play a role in one’s sense of responsibility, but it does not serve to make my actions objective to a preconceived principle, as a principal does in the sense of accountability.

The relationship between the holder and the one held that makes accountability – let us refer to them as principal and agent for the sake of simplicity – is entirely bound by an instrumental scheme. The principal wants beneficial actions done by the agent, who serves as a means to achievement. This arrangement of expectations is the core of the relationship between the principal and the agent. The principal is a principal insofar as she has expectations for an agent. The agent is an agent insofar as he acts under the oversight of the principal. This is a relationship in only a superficial sense. Each party does what serves himself/herself under the constraint of what the other desires or wants. There is no expectation of sincere consideration for the other in this relational schema, the absence of which minimizes, even eliminates, the relational sense of it. In other words, the moral or custom is dictated by a fully inward consideration of one’s objectives. This collapses any real sense of morality, a feature that frequently plagues institutions and organizations premised on impersonal relationships.

What do I mean by an impersonal relationship? Simply put, it is one that is dictated by formal rules. The rules are not suggestions. They are prescriptions, stipulations on a course of action. The challenge for the principal is not to express the rules of action; the challenge for the principal is guaranteeing that the rules are followed. The agent can have been prescribed how to act, can be fully aware of the rules. But even then, the agent may not perform. It is at this point that accountability is thought to function. Accountability is imposed on the agent. It is an instrument of the principal. In other words, it is an instrument of power over another where formal rules determine how another must act. The starting point of any accountability relationship, then, is the objective of the principal. A perfect accountability scheme (we will revisit this idea of “perfect” below) requires that the agent perform a prescribed role. To the greatest extent possible, the complications involved in performing this prescribed role have been addressed through the formal rule scheme.

One way of thinking through this impersonal relationship is by seeing it as a contest over the decision. The agent is in a position through which his decision will have ramifications for the principal. The principal, meanwhile, wants to minimize the decision-making scope of the agent and enforce arrangements that accord with her decision (of what is to be done). A perfection of accountability is a condition of nondecision for the agent. This is one reason why substantive accountability is conceptually convoluted: procedural accountability designs away the substance of the decision itself.

The condition of nondecision is one facet of accountability. It is the condition that Dubnick and Frederickson refer to as pre factum accountability, or accountability before the deed.⁴ They contrast this facet of accountability with post factum accountability, which places focus upon consequences as a facet of the principal–agent relationship. It is reasonable to see the necessity of consequences as made by a failure of control. When the condition of nondecision in pre factum accountability fails, the force of consequence in post factum accountability arises.

2.3 Reconsidering Per Factum Accountability

In a previous essay,⁵ I added to the dual concepts of pre factum and post factum accountability the idea of per factum, which I described as during or through the deed itself. Viewed through administrative logic, in particular the solutionism of technics, per factum accountability is better thought of in the sense of “thoroughly done,” the sense bearing a connotation of perfection. Achieving the condition of nondecision through thorough design is perfect accountability. But the political consideration eschews perfection in that sense. The focus must be on the deed itself, the ongoing process of any activity that is supported by authority and involves power. In this respect, the per factum accountability should be viewed as ongoing, as being “through” the deed, in the same sense that “permeate” means to go through and “permit” means to send through. This meaning of per factum does not correlate with a sense of perfection, though, because of the contestable practices intrinsic to politics. Simply put, the administrative logic of accountability begins and ideally ends with pre factum, meaning a properly and thoroughly designed system or institution that minimizes the concern over the execution of the deed (per factum) and the possibility of incompletion (post factum). While this may itself appear impossible, it is the basis of accountability design: a system that functions impersonally does so without judgment or thought of its performing elements (in an organization’s case, people are the elements). In other words, perfection renders moot the question of performance, and thus the per factum and post factum concerns are irrelevant.

We should not be distracted by the challenge and the fact that such design is not possible in many cases. Particularities will always make such “perfection” inconsistent with practice. But inconsistency is not inconceivability, and the point is that every error in the system is perceived as an opportunity for improving the pre factum system that is behind it. This is an ethos of technology. The question should never be posed in terms of a given state, a fixed condition, because at issue is the condition of perfectibility, not the state of perfection. The administrative sense that is behind solutionism is supported by a belief in a perfectible state or institution.⁶ This belief is not hindered by the evidence of imperfection since that is considered evidence for what can be improved. Such is the logic of technics, and it is thus that accountability is realized as a concept of technology.

The idea that accountability is possible “during” the deed, as I imply with the sense of per factum, is conceptually inconsistent with how accountability is generally conceived within administrative logics. In other words, opening the deed is a direct challenge to accountability and exposes the problems with it. These problems are largely the result of accountability being a concept of and for technology, a concept in diametrical opposition to responsibility.

It is worth returning to the question of responsibility to clarify the claim that accountability is in both concept and practice against responsibility. The distinction hinges primarily on how accountability is premised on impersonal structures and institutions to facilitate preconceived actions. Responsibility, however, operates within ambiguity and uncertainty, conditions that require judgment about how to act in particular circumstances. As already mentioned, it is understandable to speak of “my responsibility” because one must take possession of the decision that corresponds with the action directed towards the particular concern. However, the grammar of accountability does not allow us to speak of “my accountability.” This phrase is senseless. I can say, for example, “Raising my children is my responsibility.” This connotes a relational understanding of my role, what we can call a caretaker, and its meaning is built upon this relationship. The actions that correspond with “raising my children” are not preconceived, even if I hold firmly to certain standards, morals, customs, or even principles. The point is that my responsibility may require me to act in a particular case against such standards and then to answer for doing so (hence the account within the concept of responsibility).

I am unable to say “Raising my children is my accountability” and to convey a sensible meaning. The reason is that accountability produces an object relationship so that the “I” that speaks cannot express something about the “I” through accountability, but only about the tasks and expected behaviors. To speak in this way of accountability requires me to say “I am accountable for raising my children,” a remark that connotes a person or structure that “holds” me accountable and “possesses” my condition of accountability. The “I” then turns to the tasks necessary to fulfill the accountability relationship. What accountability enables above responsibility is an objective view on what is necessary to be done, objective to the extent that another party is charged with determining the appropriateness of the act based upon established standards. It is this last qualification that distinguishes “being held accountable” from “my responsibility,” and this distinction is further clarified by the difference between action and tasks.

2.4 Tasks and Design

Consider what is connoted by task: a duty, a chore, an assignment, a charge. A task is something given to you based upon what is expected of you. It is often generalized by the expected conditions that demand activity. I want to contrast this with the idea of action, which I believe connotes a greater degree of ambiguity sufficient to support the judgment of the actor. A task is itself an idea that eschews judgment, while action demands it. The distinction is clarified by Arendt’s description of action as being the conditio per quam (the condition by means) of political life, a category she contrasted with work (the category of artificial worldliness) and labor (the category of life in the sense of the biological).⁷ What is specific to action is what Arendt called natality. Action is the condition by which a person can start something new, can bring about a beginning, and it is partly for this that action is a condition of freedom itself. Natality is first realized in the fact of our birth, that every birth is a new beginning, and our capacity to create something new is what makes us free.

Arendt’s conception of action is diametrically opposed to the conditions that are fostered through the bureaucratic apparatus through which accountability functions. The actualization of freedom through action is the capacity to do the unexpected, which is how a new beginning is brought about. “It is in the nature of beginning that something new is started which cannot be expected from whatever may have happened before.”⁸ Routine, standards, and procedure all work against the conditions of action precisely so that expectations are realized, so that the unexpected does not happen.

But accountability and the associated administrative logics are not to be found in the other “human conditions” that Arendt discusses. There is no biological need for accountability, so it is not present in labor. Work, according to Arendt, is judged based upon the production of a world suitable for human use. It is associated primarily with the production of the material world, the kind of artifice that we associate with the Promethean violence against nature,⁹ wresting from it the materials for our needs and goals. Work characterizes the condition of making the material world accommodate human life. Accountability is about task, not work. It is not premised on human use but on the use of humans. Accountability is complicit in the bureaucratic arrangement that prioritizes artificial production above human use, the quality of dehumanization that is especially associated with bureaucracy.

It does not have to be this way. There was a sense of accountability that was associated with representation and a liberal sense of government, an understanding that derived accountability from responsibility. But that idea no longer prevails in the meaning of accountability. The modern idea of accountability rests on notions of control over the administrative apparatus to ensure execution. It is a concept that follows the growth of bureaucracy and administration to such an extent that responsibility is conditioned by accountability. Action, as Arendt described it, is consistent with responsibility but not accountability. Accountability is a political concept only to the extent that it domesticates the components of politics to the point of erasure. What is done through accountability is not a space of contestation where the possibility of action might produce the unexpected; rather, what is done is a restricted space of behavior, where ultimately an algorithmic expression is possible, thus erasing action and responsibility. Behavior is the antithesis of action. It is produced by design, and behavior is to act in a way consistent with expectations.¹⁰ Design implies expectation, and accountability is a concept of design insofar as the agent is given the procedures and the incentives to do as the design dictates. Actors properly embedded within an accountable institution do not act where their tasks are not prescribed, which is a way of saying that action is conceptually inconsistent with accountability itself. Prescription is the arena of behavior and the act of constraining action; or, more concisely, prescription erases action. Therefore, one can never say “my accountability”; the prescription of one’s actions removes any ownership over the act itself.

The early debates over accountability were focused on what was called “administrative responsibility.” The sense of administrative responsibility was eventually abandoned in favor of talking about accountability, and I argue elsewhere that this shift in language testifies to a changing idea of the meaning of “responsible government.”¹¹ Carl Friedrich, who appeared to take the side of administrative discretion and deference to expertise, stated that

if a responsible person is one who is answerable for his acts to some other person or body, who has to give an account of his doings (Oxford English Dictionary), it should be clear without further argument that there must be some agreement between such a responsible agent and his principal concerning the action in hand or at least the end to be achieved. When one considers the complexity of modern governmental activities, it is at once evident that such agreement can only be partial and incomplete, no matter who is involved.¹²

Friedrich’s apparent defense of discretion in fact opened the door for design. He goes on to say immediately after this remark that a realistic consideration of the electorate and its representatives shows that, as principals, they are inadequate to the task of reaching such an agreement over administrative tasks. They are unable to bring about what Friedrich calls “responsible conduct of public affairs,” by which he meant the alignment of required actions and defined ends, with one qualification: “unless elaborate techniques make explicit what purposes and activities are involved in all the many different phases of public policy.” Indeed, the qualities associated with modern accountability – performance measurement, transparency, incentives of various sorts, rules, and procedures – are the elaborate techniques that fulfill this qualification. These techniques render moot the story from the “responsible person” since they address the problem from a design perspective, meaning the “many different phases of public policy” are considered in dictating how an agent acts under relevant conditions.

A fair rejoinder would point to the way that Friedrich himself defined responsible: one who is answerable for his acts to some other person or body, who must give an account of his doings. He cited the Oxford English Dictionary. The current version (OED Third Edition, March 2010; most recently modified version published online June 2021) offers the following as the first definition of responsible: “Capable of fulfilling an obligation or duty; reliable, trustworthy, sensible.” The third, fourth, fifth, and sixth definitions under the adjectival use refer to some variation of answering a charge (even using “accountable for” and “to be held as”). There is no dispute here that responsibility includes some facet of account-giving, but the distinction between accountability and responsibility is the extent to which the account remains a story communicated between two people about what happened. Modern accountability takes the story as a component of control, as though the story must be written beforehand. Responsibility entails an account about why one decided to act as one did, which in some respect is the arena of “substantive accountability.” In perfect form, though, accountability takes the account and makes it an ex ante concern that determines what is to be done. At issue is how accountability takes the component of the account and magnifies it. In responsibility, I may feel obligated to explain to someone why I did what I did. With accountability, that account is transformed so that I act as you intend, making the story itself not one that the account-giver tells but rather a story to be followed.

The emphasis on the account that coincides with modern accountability is not about the act of explaining one’s action; the emphasis is on controlling the account itself, on dictating the story. To refer to accountability is not to refer to a human quality but to a technological quality in which systems are designed to guarantee outcomes. The “ability” in accountability is a quality of the institution or system, not the person. When we refer to an accountable actor, we refer not to a person capable of action and responsibility but instead to an agent performing within a designed institution or system. The perfection is aimed at the task, and the person is an instrument within a perfectible system. One could say that accountability seeks to solve the problem of responsibility by perfecting it. Responsibility conceptually retains the possibility that a person might act on their own discretion and judgment, and accountability removes this by and through design. It is not that accountability is different from responsibility. Rather, it is an extreme form of it expressed through the administrative logics of solution and technology. Referring back to the definition of responsibility above (capable of fulfilling an obligation or duty; reliable, trustworthy, sensible), the -able of accountability erases the uncertainty of the second part of the definition. The design approach seeks to exclude the variable effect of reliability, trustworthiness, and sensibility. The one holding the agent accountable is an institution.

2.5 Perfecting Bureaucracy

Let us return to the question of perfection in accountability. The perfection of accountability rests in the same ideal condition of perfection identified more than a century ago in the bureaucratic system of order described by Weber: in dehumanization.¹³ He defined this quality of the bureaucratic apparatus as being composed of agents acting sine ira et studio, a phrase he might have borrowed from Tacitus, meaning without anger or bias. Tacitus meant by this that a story of historical occurrence (an account) might be presented to an audience based solely on facts of what happened and without any flourishing from the speaker. In this way, an orator is restrained from attempting to influence the audience. In essence, Tacitus refers to what we might consider an “objective viewpoint” and used this formulation in his early history of the Roman Empire. Weber’s meaning is slightly different. His meaning refers to the acts of agents in a bureaucratic system. It is the deed, not the story, that concerns Weber’s notion of objectivity behind sine ira et studio. Nevertheless, both Tacitus and Weber refer to a condition that is restricted by a conception of objectivity. Both sought to introduce a condition through which fact (that which is done) supersedes truth to such an extent that fact itself becomes truth. But Weber’s description did something that Tacitus’s did not. Tacitus’s conception of an objective account allowed interpretation by the listener. A story can be told “faithful to the facts” but the meaning can remain open to the audience. Tacitus’s sine ira et studio meant that the account must be given according to what was done. There thus remained some political element concerning how what was done could be relevant to current matters of action and deed. The sine ira et studio expressed by Weber is not concerned with what was done but rather what is to be done. Meaning, interpretation, and understanding are expressly removed from the question of fact, understood to mean deed or that which is done. The agent of the Weberian sine ira et studio is foremost an instrument of the design. This is the meaning of dehumanization: a prescription to ensure that design becomes fact. This quality is of utmost importance to bureaucracy and administration. Weber called it the special virtue in his description of the ideal-type bureaucracy.

Weber answered the issue of what is to be done by upholding the function of calculability in the deed itself. Here is how Weber expressed it:

The peculiarity of modern culture, and specifically of its technical and economic basis, demands this very “calculability” of results. When fully developed, bureaucracy also stands, in a specific sense, under the principle of sine ira ac studio. Bureaucracy develops the more perfectly, the more it is “dehumanized,” the more completely it succeeds in eliminating from official business love, hatred, and all purely personal, irrational, and emotional elements which escape calculation. This is appraised as its special virtue by capitalism.¹⁴

Perfection (fully developed or thoroughly done) requires dehumanization. This is the mode through which the bureaucratic apparatus exerts control, and it is precisely the condition of modern accountability. What is to be done is predetermined so that actors behave according to tasks. The story is told beforehand. This is markedly different from the sense of responsibility, which retains some sense of moral conduct by the individual, a conduct that depends upon the exercise of the faculty of judgment. Accountability operates where judgment is made obsolete.

Accountability is the moral expression of amoral conduct that marries the conditions of perfection in bureaucracy: dehumanization and calculability. How can one claim accountability to be a moral expression of amorality? The simple answer to this is that a moral expression is concerned with the central principles of right and wrong as pertains to right conduct; amoral conduct is carried out without regard to right or wrong. This is precisely the condition of calculability (morality as determined by calculable rules of conduct) and dehumanization (disregard for personal considerations of right and wrong). How does accountability marry these conditions together? A necessary premise of accountability is that the actor behaves in accordance with established rules and procedures and not according to their own judgment of appropriate conduct. In other words, “Objective discharge of business primarily means a discharge of business according to calculable rules and ‘without regard for persons.’”¹⁵

2.6 Contestation Is Not a Fix

Modern accountability differs from responsibility. The key distinction is that accountability is conceptually linked to the use of humans as part of an instrumental institutional apparatus, which makes it an extreme version of certain components of responsibility. The apparatus, which in modern parlance is called bureaucracy, puts primary emphasis on the tasks and procedures for achieving defined goals. We may speak of an accountable person, but in effect we are describing the system in which the person operates. Persons cannot have accountability because in being accountable they relinquish action to the operations of tasks and thus to a power that holds them accountable. Accountability operates as a condition of perfectibility for bureaucracy.

Accountability cannot be political insofar as it is premised on the erasure of politics from the discharge of state operations. Conceiving of accountability as political has sparked interest among many scholars (the present author included), who hold that accountability is an essential precondition for democracy and that there is the possibility of democratic accountability in the modern sense of government. But there is not. Rather, accountability operates against democracy in the important sense that it perfects bureaucracy, not democracy. What I previously identified as per factum accountability – something between pre factum and post factum that took place during the deed – suffers the same problem as post factum accountability: everything depends on the perfection before the deed. Accountability operates against the unexpected; its functions seek fulfillment of expected outcomes through increasingly optimal designs. The unexpected of today becomes a bit of new information for the greater perfection of the system. That is what a thoroughly done (per factum) bureaucratic system is in practice. Accountability is a concept of and for technology and design, not a concept for human use but, at the risk of repetition, it is a concept for the use of humans. While it is the case that in its early stages humans (principals) use humans (agents), the logic persists in such a way that rules and algorithms ultimately prevail. In other words, the design is intended so that eventually the tasks of the agent are compelled by evidence.

This logic is one of the reasons that rethinking accountability by embedding contestation or by emphasizing a substantive orientation of account-giving fixes nothing. Modern accountability entails practices that absorb what was not used where and when it is useful to do so. The shift in focus that occurred with modern accountability was not a move away from representation but rather a step further into the direction set by representation where the state addresses the increasingly complicated and specialized conditions of modern culture that derive from the interests of the people. Accountability is its own solution. When things go wrong in modern mass society, the calls for accountability ring out almost immediately. Sometimes this means punishing the wrongdoers, but it is also a call for preventing it from happening again. What this ultimately looks like is no surprise: refinement of rules and procedures. Contestation is a value similar to transparency that can easily be normalized under the broader goals of accountability. This is in practice quite common. Public participation and transparency are both facets of modern accountability arrangements. The point is not that they are present in the arrangement. The important point is that they are subsumed under accountability itself.

The problem in accountability has always been who is in control. This is apparent from the earliest stages of the debate. It is at the very heart of the dispute between Carl Friedrich and Herman Finer: is a responsible administration better guided by experts or public representatives? The entry of “the public” into this debate is not a progressive step, nor is it necessarily a normative good.¹⁶ Increasing the publicness of an accountability regime or arrangement does not address the issue of control. What it does, however, is to highlight that the control falls to nobody. The public is not a somebody but a calculated political entity designated by its generality. It fits the essence of calculability (in, e.g. “public opinion”), the quality that provides bureaucracy its technical superiority: decisions are based on measurable criteria and not left to somebody. The decision must be for the “public benefit.” The decision is what accountability addresses, and the modern form of it eschews representation in favor of objectivity, an objectivity that is realized in nobody deciding. Questions about the exception and the unexpected, the questions that constitute political concerns, are erased by design.¹⁷

An agent under an accountability regime asked to justify or defend a decision presents two problems. The first is that it is not the agent’s decision. This is not meant in the banal sense that the agent is simply following orders from a principal because it is rarely so simple, although this is an issue of substantive accountability and is the source of perversions of accountability, such as scapegoating. More critically, though, the decision is increasingly determined by the measurable components of the substantive policy arena. The orders come from nobody. They can be based possibly on evidence, data or information. The second issue is to whom the decision must be defended or justified. The defense here must balance a finely tuned distinction between justification and manipulation through what can only be properly called propaganda.

For example, let us imagine an official in a national health department, such as a leading infectious disease expert. This expert identifies a highly infectious virus that leads to greater than average mortality. She briefs government officials and explains that certain difficult measures must be implemented in order to stem the spread of the disease. Cities shut down commerce, schools close, and only essential services are allowed to continue operating. The expert is then scheduled to speak on dozens of news programs to explain what she sees in the data and to exert her expertise in defending the judgment to take such drastic measures. She explains that this will save lives and that the sacrifices are necessary to stem the spread of the virus. She explains the decision with admirable clarity and cites data and evidence. But most people cannot understand the evidence and cannot interpret the data. They express doubt. They contradict the expert with their own data and evidence. They question the integrity of the expert. At some point, the attention turns from justification to winning the argument. The expert is joined by other experts to spread a message and to counter “home remedies” and “alternative facts.” Her status as an expert is questioned by her opponents, who claim that she is nothing but an opposition bully trying to undermine the people and destabilize society. Meanwhile, commercials and billboards and public service announcements encourage people to take the steps necessary to stem the spread of the virus: avoid socializing, wear a mask, and wash your hands.

One could tell a similar story about a monetary expert who faces a potential catastrophic financial collapse. During the 2008 financial crisis, the US Fed began a policy of quantitative easing, a monetary policy that had to that point been scarcely used in a few countries, including Japan in the early 2000s. The experts in the Federal Reserve Bank faced the challenge of explaining a complex purchasing practice that was intended to encourage banks to protect their balance sheets enough that they could make loans and, thus, send money into the economy. To nonexperts, this looked like the equivalent of printing money, which sparked fears of inflation, but the experts had to make nuanced arguments on the differences between money printing and quantitative easing. A little more than a decade later, facing another potential financial crisis from a global pandemic, the US Fed began purchasing Treasury bills again, but this time the leader of the Fed was adamant that the practice was not quantitative easing. Instead, the practice was framed as a way for the Fed to create reserves. They purchased short-term rather than long-term bills and were not making the purchases with the goal of increasing liquidity. As then Fed Chairman Jerome Powell said, “This is not QE. In no sense is this QE.”¹⁸

Both descriptions of experts at work highlight the disparity of knowledge between the experts and the public at large. The expert health official case could be seen as an example of contestation, and it could be viewed as an example of substantive accountability. The expert is put in a position of justifying the activities (policies) that she deems necessary based on the evidence. On its face, this would seem to be an exercise of accountability. But, I think that it actually exemplifies a failure of accountability as I understand it. The example displays the justification of accountability as necessitated by the increasingly complicated conditions of modern government. To grasp the failure, consider the way that Mill described accountability:

When the accountability is perfect, the interest of rulers approximates more and more to identity with that of the people, in proportion as the people are more enlightened. The identity would be perfect, only if the people were so wise, that it should no longer be practicable to employ deceit as an instrument of government; a point of advancement only one stage below that at which they could do without government altogether; at least, without force, and penal sanctions, not (of course) without guidance and organized co-operation.¹⁹

Mill alluded to the ingredients necessary for going from representation to administration. Perfection approaches the point when the people could do without government, by which he seemed to mean the use of force and penalty. There will always be the need, one infers, for guidance and organization. The perfection of accountability in the liberal sense makes the purpose and function of representation unnecessary. If the interest of the representatives approximates the interests that identify the people, then the people do not need representation. This only works, though, to the extent (in proportion with) the enlightenment of the people. Mill’s liberalism depended upon expertise and leadership of those who were knowledgeable, as he made clear in his advocacy for weighing voting rights based on education. For Mill, this was a solution to the problem of aristocratic power. He believed that knowledgeable voters would improve upon the political decisions made by property owners. By the early twentieth century (and into the twenty-first), this same argument of giving more political power to those who know would be used to tame democracy (which of course was the original target of this argument, dating back to ancient Greece). The perfection that Mill described in the nineteenth century had a different nuance by the middle of the twentieth century. The person who is accountable is not a representative but an administrator. They are not supposed to have interests (sine ira et studio) and are supposed to act according to calculability (measurable, evidence-based reasoning). The point at which an administrator must explain to the public the reasons for a policy decision is a failure of accountability precisely because the decision is under question. This activity is aimed at bringing the interests of the public into line with what the expert has deemed appropriate based upon evidence and data. As Mill himself described it, the perfection of accountability entails improving the wisdom of the public (since, ab definitio, the expert knows). When the public displays an ignorance about the issue at hand, we witness the imperfection of accountability.

This standpoint is clearer in the context of administrative accountability because, as described above, the – able of accountability rests in the institution, not an individual. At issue is the story about what is to be done, as in telling the story ahead of time. In other words, the ability of accountability situates power into the nobody of administration (public administration), and the expert is the remnant agency of this nobody. That the public is insufficient for the task of governing is a key premise of public administration. Woodrow Wilson regarded public opinion as being meddlesome and considered it the role of the reformer to “persuade [his fellow-citizens] to want the particular change he [the reformer] wants.”²⁰ Walter Lippmann said that “the false ideal of democracy can lead only to disillusionment and to meddlesome tyranny.… The public must be put in its place, so that it may exercise its own powers.”²¹ Finally, consider Carl Friedrich, in his explication of what he called administrative responsibility:

The pious formulas about the will of the people are all very well, but when it comes to these issues of social maladjustment the popular will has little content, except the desire to see such maladjustments removed.… Consequently, the responsible administrator is one who is responsive to these two dominant factors: technical knowledge and popular sentiment. Any policy which violates either standard or which fails to crystallize in spite of their urgent imperatives, renders the official responsible for it liable to the charge of irresponsible conduct.²²

What Friedrich, Wilson, and Lippmann are describing is the arrangement for a technical civilization that, while being sensitive to public opinion or sentiment, also must, as Wilson put it, “make public opinion efficient without suffering it to be meddlesome.”²³ And as Friedrich argued, this new kind of responsibility, which I have argued is modern accountability, must serve the double standard of adhering to technical knowledge and remaining sensitive to public opinion. These are the requirements that Friedrich described as a novel type of responsibility for the permanent administrator that we now call accountability.²⁴

So how does this story about the health administrator indicate a failure of accountability rather than an exercise of it? The short answer is that it only fulfilled one of the two standards: it failed on the popular sentiment front. The example illustrates also how accountability operates ideally as a pre factum course of action: the tasks must be seen as necessary according to the evidence and information, or according to the technical details of the issue. Accountability is an operation that is primarily concerned with the decision, and in its perfect state, the decision is preordained by necessity. Popular sentiment must be “accounted for” in the decision itself. If it reaches the point of persuasion (force), then it is imperfect. Thus, a key part of public accountability is constructing what Mill called the identity of the people.

For Mill, this perfection of accountability centered upon the public sentiment merging with representative interests. The new sense of responsibility arising in the twentieth century that Friedrich remarked upon is not representative in nature but rather centered upon expertise and public sentiment. The difference is not negligible since, assuming the expert acts in good faith to her profession (a key proviso in Friedrich’s model), the contest will inevitably favor the expert. Contestation in this regard contradicts the premise of expertise. Mill’s notion of perfect accountability fits the administrative conception of accountability to the extent that the perfection of accountability requires an alignment of what is to be done with the enlightenment of the people – in other words, the people are in no position to dispute the necessity of evidence. The substantive accountability in which justification is required is an act of persuasion, but it may as well be an act of manipulation. The expert knows. The public doesn’t. The expert has to convince the public. This is the justification. If the public disputes the expert, then the very premise of the expert’s position flounders. Thus, contestation is not a way of improving or fixing accountability. It is an act of destruction. The contest calls into question the very legitimacy of the institutional arrangement by putting into question the premise of expertise. The Millian standpoint retains an element of the public sentiment that Friedrich mentioned, but the administrative concept of accountability incorporates public sentiment as another component of perfection, a problem to be solved. To contest the decision within an accountability framework is a destructive act, but as fits the logic of technology, destruction is constructive. There is the ongoing possibility of subsuming the content of the contest for the sake of improving accountability, to the extent that the contest reveals improvement or the possibility thereof. The accountable institution can incorporate the values or ideas into the accountability framework, whether that be through formalizing contestation in public participation or transparency or by using the problems raised in the contest to improve the impersonal operations of the accountable institution. Herein lies the perfection of accountability, a thoroughly done task achieved by design that will further alienate the nonexpert from the decision itself. The realization of per factum components of accountability deepens bureaucracy.