TOC Previous Next A+A-Print


Question 193: Should those in authority seek consensus or make decisions unilaterally?

I am a high-ranking civil service employee in a department of the U.S. government. The managers above me are political appointees who come and go with changes in the administration and sometimes between times. In the years I’ve worked here, I’ve experienced different management styles, each with its own problems.

At one extreme is the autocratic style. The political appointees talk almost exclusively to one another and make the decisions, and we underlings are expected to implement them without arguing. This inevitably leads to many ill-informed decisions due to lack of experience, and often such decisions are virtually impossible to carry out and cause a great deal of frustration and foot dragging on the part of permanent employees, including me.

At the other extreme is the democratic style. The political appointees set up numerous committees and hold endless meetings in an effort to build consensus. This is time consuming, and, due to deep-rooted philosophical differences and the complexity of many of the issues, it seldom produces the desired consensus. Instead, it almost always divides the permanent employees—and sometimes the political appointees themselves—into factions. The result sometimes is that nothing is done and problems remain unaddressed. Or else, to end the deadlock, decisions finally are made at the top level, and, while these are likely to be well-informed, the losers are frustrated.

We are now about to undergo yet another transition. Surprisingly, the person who has been nominated (and undoubtedly will be confirmed) as secretary of the department came by and asked everyone at my level to prepare a brief statement on the main challenges the department faces and what should be done to deal with them. I am going to seize the opportunity to address the overarching problem of management style. Of course, I already have my own thoughts on the matter, but I am seeking input from others as well.

Perhaps you will take this to be a problem in management technique rather than a moral question. However, I will be grateful for whatever you have to say from the moral point of view.


This query is both a technical problem and a difficult moral question: How should authority be exercised? The answer: Diverse styles are appropriate in diverse situations. More precisely, then, the question is: What style should the department’s managers prefer in making decisions to guide the work of employees at the questioner’s level and below? This question calls for the derivation of specific norms. Since authority should be exercised by those in the best position to make sound decisions, the managers should not usurp their subordinates’ proper authority. In directing others’ work, the managers should prefer a consultative style of decision making, by which all who will carry out the decision are encouraged to participate in deliberation; but, unless consensus spontaneously emerges, the managers alone should make the choice.

The reply could be along the following lines:

Situations that call for a decision and in which some of the proposed options are immoral involve considerations that would complicate discussion of your question. I shall consider only cases in which all the options under consideration are morally acceptable, so that employees could cooperate in good conscience with any decision the political appointees made.

The primary responsibility of those in authority is to make decisions that shape cooperation toward realizing the group’s common purposes. Failure to proceed rightly in making decisions, as you point out, leads to bad decisions, undermines cooperation, and prevents the group from achieving its ends, with the further bad result that its most dedicated and conscientious members remain unfulfilled, become frustrated, and lose the motivation to do their work, so that they either resign or, as you put it, drag their feet. Therefore, your query—How should managers make decisions?—is both a problem in management technique and a difficult and important moral question. In responding to the moral question, I shall articulate some norms that I believe have wider validity in a way that will specify them to the situation you have described.

Plainly, diverse styles are appropriate in making diverse sorts of decisions. In emergencies, for instance, autocratic decision making can be indispensable; in matters that greatly affect employees’ morale but have little other impact on their work’s quality and productivity, decision making ought to be democratic. Moreover, competent managers often delegate their authority to others. But I assume you take for granted this variety of decision-making styles, and so I understand your question as more specific: How should your department’s managers prefer to make decisions directing the work of employees at your level and below?420 I shall cover the points I think you should make regarding that specific question.

The first point is that the political appointees, who are the top managers, need not make or even be involved in every decision in the department. Insofar as possible, they should focus their attention on a few matters: innovations and reforms in programs and policies affecting the work of the department as a whole or its major sections, internal conflicts not resolved at lower levels, and major external threats to the department’s integrity or work. True, nothing in the department can be excluded from the purview of the political appointees, and they should intervene whenever necessary to correct serious abuses. Still, they normally should not attempt to micromanage by making day-to-day decisions directing routine activities, the inner workings of each office, or even the work of employees at your level in implementing established programs and policies.

The underlying principle of this distribution of decision making is that authority always should be exercised by those in the best position to make sound decisions. Carrying the principle through to its logical conclusion will mean that virtually everyone in the department will participate in authority, making some decisions that guide cooperation with at least one or a few other employees, since even workers at the lowest levels usually are in a better position to make certain decisions shaping moment-to-moment cooperation with their co-workers and superiors than are the latter themselves.

Turning to the decisions which the political appointees appropriately make, I come to a second point. Their preferred style of making decisions should be neither autocratic nor democratic but consultative. Decision making has two parts, which should be distinguished: (1) deliberation to discover and assess all available options and (2) the choice of one option to the exclusion of others. All who are able to contribute to (1)—deliberating—should be invited to share in it; that usually includes everyone who will be involved in carrying out the decision. Occasionally, well-conducted deliberation leads to consensus, a common preference for one option. Often, however, no consensus emerges, since two or more options are supported by weighty reasons, grounded in noncommensurable prospective advantages and disadvantages. As you remark, such issues often are complex and opposing positions sometimes reflect deep-rooted differences. In such cases, the individual or small group responsible for the cooperative effort as a whole must proceed to (2)—making a choice. The implications for consultative decision making of the distinction between deliberation and choice can be stated, although without literal accuracy, in terms of the two styles you distinguish: deliberation should be democratic, but when no consensus emerges choice must be autocratic.

The third point is that everyone should be free to suggest problems for consideration and those who will make the choices should consider every such suggestion, but they alone should initiate the deliberative process and regulate it. If others initiate deliberation, it will not involve the group as a whole; subgroups will pursue their own agendas, and time will be wasted discussing matters those in charge are not prepared to deal with. Those regulating deliberation should try to define the problem and seek the help of everyone involved in learning two things: the options for resolving it that are worth examining and the considerations for and against each option.

The fourth point is that, where it is appropriate for the political appointees to make a decision, they should make it clear to everyone that they mean to do so. If a consensus emerges, they of course should welcome it, but they should not expect universal agreement or waste time and effort seeking it. Head counting—that is, any form of consultative voting for and against options in order to determine a recommendation by those who do not have the choice to make—should be discouraged. It concedes the decision-making role to those whose responsibility should be limited to contributing to deliberation, and so confuses the two. Instead, the political appointees should ask for well-prepared briefs for and against each option. They also should direct that these briefs be studied and criticized by those who disagree with their contents and then rewritten in the light of all the criticisms received. In using these briefs, the political appointees should focus their attention—and make it clear to everyone that they are doing so—on the quality of the evidence and arguments, regardless of the numbers and status of the permanent employees supporting particular conclusions.

Someone might object that numbers and status can hardly be ignored, since the wholehearted support of those who will implement a policy is important to its success. While the support of subordinates who will implement decisions is necessary for their success, however, subordinates should support any legitimate decision, and those in authority should enlist their support without abdicating their own responsibility for making decisions. Of course, given pervasive habits of insubordination and other sinful social structures, leaders must take into consideration as a disadvantage any residual probability that subordinates’ halfheartedness in implementing a decision will limit or negate its potential benefits. But that must not be confused with information about the subordinates’ views on options. Many loyal and dedicated employees will carefully and energetically implement a decision for an option other than the one they preferred.

The fifth point is that in regulating deliberation the political appointees should review its results as they come in and regularly consider reformulating the problem, recasting the alternatives under consideration, widening the deliberative process, and so on. This is necessary to overcome the limitations of initial ignorance and take advantage of emerging data and insights. Moreover, only constant monitoring can lead to necessary procedural choices, such as to suspend deliberation by setting the problem aside for the time being or to terminate it with a definite decision and the beginning of its implementation.

The sixth point is that, when decisions are announced, the reasoning behind them almost always should be articulated, since, by clarifying the purpose in view, the reasons for a decision contribute to its accurate interpretation and provide the intrinsic reasons for implementing it. Oftentimes, the announcement of a decision also can enlist the support of those who favored some alternative by indicating, at least briefly, how the considerations supporting alternatives will be taken into account in other phases of policy making.

The final point is that this consultative process will work well only if everyone involved takes pains to communicate precisely and clearly but briefly. If those initiating and regulating the process fail to communicate carefully, other participants will not know what is expected of them and probably will develop false expectations about their roles in the process. If participants who try to offer their information and ideas are not clear and precise, or if they are verbose, their potential contribution will be, practically speaking, more or less unavailable to decision makers.

One can hope that, if this procedure is followed, the disadvantages of both the so-called autocratic and democratic styles will be avoided. The political appointees will talk with and hear from everyone with anything to contribute; their decisions will be well informed; and the reasons for them will be articulated as well as possible by those who favored the option selected, taking into account others’ criticisms. While not everyone will agree with the final decision, implementing it will be possible and more likely to succeed. When consensus does not emerge spontaneously, time and effort will not be wasted seeking it. The clear awareness on all sides that certain choices will be made at the top level will incline employees at your level not to anticipate the outcome by taking an overly firm position. Excluding head counting and focusing on the quality of evidence and arguments for positions will discourage the formation of factions and instead promote more effective gathering of relevant data, more thorough analysis and criticism, and more willing acceptance of decisions. Finally, there will be less of something that undermines sound deliberation and is common to both the autocratic and the democratic styles—I mean the tendency of subordinates to tell superiors what they seem to want to hear.

Like other problems in interpersonal relationships, inadequacies in management styles such as you describe seldom are entirely the fault of those in authority. Subordinates’ wrong attitudes and false expectations often impede sound decision making. Political appointees, for example, rightly take into account the considerations that motivated voters to favor their party, and the decisions they make based on such considerations should be respected by civil servants even if they personally favor the opposing party’s policies. When a new administration must deal with various groups of uncooperative subordinates, distortions in the management process are inevitable. Therefore, while I have not focused on the duty of those at your level to obey the political appointees to whom you are subordinate, I suggest you reflect on that responsibility and also encourage your colleagues to examine their consciences regarding it.

420. The question does not concern how the managers as a group should make decisions, and so the response does not deal with that. In their case, however, a collegial process leading to consensus might well be appropriate. See Thomas A. Baylis, Governing by Committee: Collegial Leadership in Advanced Societies (Albany, N.Y.: State University of New York Press, 1989).