4
Who, Me?—The Bureaucrats Look at the Issue of Democratic Control

Bureaucratic decision making creates a challenge for democratic political institutions because the world of democracy and the world of bureaucracy clash in a critical way. Democracy is premised on self-governance, whereas bureaucracies are populated by appointed officials whose power derives from their expertise and institutional position, not from the mandate of the electorate. Until now my vantage point has been that of the world of democracy. I have examined the democratic norms implicit in various control strategies, and I have explored the costs to the would-be controller of exerting various forms of control. As long as the issue of democratic control of bureaucracy is viewed solely through the eyes of the controller, however, our understanding will be incomplete. It will ignore insights derived from the other world involved in the clash: the world of bureaucracy.

In the next three chapters I shift my vantage point from the world of democracy to the world of bureaucracy. These three chapters introduce two major complications to the map I have developed of strategies for democratic control. The first complication consists of the addition of

another perspective to that of the would-be controller—the perspective of the bureaucrat. The second complication consists of introducing the consequences of differences among policy areas. By adding these two complications I add two more layers of complexity to the question of choice among control mechanisms. I argue that the would-be controller should consider both the attitudes of the bureaucrats and the nature of the policy area in choosing a control strategy. But while complicating the would-be controller's choice, an understanding of bureaucratic attitudes and policy area variations also suggests avenues for designing strategies of democratic control more effectively.

In this chapter I begin my examination of bureaucratic attitudes by discussing how bureaucrats justify their resistance to external control. In chapter 5 I step back from the bureaucrats themselves and discuss the ways in which the world they work in varies by policy area; I also discuss how these variations affect the would-be controller's choice of control strategy. In chapter 6 I return to the bureaucrats to examine how variations in the world of work create variations in bureaucratic perceptions of the world of democratic politics, and thus create opportunities for bureaucrats to accept some forms of control.

All strategies for controlling bureaucracies are dependent to some extent on the acquiescence of the bureaucrats involved.^[1] Opposition will at best force the would-be controller to expend extra resources on the control

[1] Herbert Kaufman, focusing primarily on internal control issues, argues that "subordinates may know precisely what is expected of them, be perfectly capable of doing it, and still not do it. What they are asked to do may offend their personal principles or their interpretation of professional ethics or their extraorganizational loyalties and commitments or their self interest…. When orders from above conflict sharply with their values, they quietly construe the orders in a way that makes them tolerable" (Kaufman, Administrative Feedback [Washington DC: Brookings Institution, 1973, pp. 3–4).

effort, at worst lead to the failure of the control effort to constrain bureaucratic behavior. Scholars of bureaucracy have long noted that far from welcoming external intervention, bureaucrats typically resist it.^[2] These scholars have generally explained the displeasure bureaucrats express at intrusions into their domain by relying on explanations rooted in the ways institutional structures provide incentives for self-serving behavior. Bureaucrats resist control because they enjoy autonomy, because their lives are easier if they are their own masters, because they feel that they know best.^[3]

There is little reason to dispute this portrait of the selfserving bureaucrat. Surely this kind of motivation lies behind bureaucratic resistance to control. But a picture that looks exclusively at bureaucracies per se is incomplete. Bureaucrats are both workers in a bureaucratically structured organization and citizens of a democracy. Thus they are subjected not only to the incentives of the agency but also to the norms of the democracy.^[4] A complete understanding

[2] James D. Thompson, for example, argues that organizations seek to buffer themselves against the outside world (Thompson, Organizations in Action [New York: McGraw-Hill, 1967], p. 20). Anthony Downs notes the strong bureaucratic preference for stasis, particularly in the face of external efforts to alter the status quo: "Bureaus have a powerful tendency to continue doing today whatever they did yesterday," he writes (Downs, Inside Bureaucracy [Boston: Little, Brown 1967], p. 195). And Robert K. Merton argues that "bureaucratic officials … resist change in established routines; at least those changes which are felt to be imposed by others" (Merton, Social Theory and Social Structure [Glencoe IL: Free Press, 1957], p. 201).

[3] Merton writes of a bureaucratic "pride of craft" that leads bureaucrats to resist change (Social Theory and Social Structure , p. 201). Hugh Heclo emphasizes a disposition to reduce "the agonies of change." (Heclo, A Government of Strangers [Washington DC: Brookings Institution, 1977], p. 143). Downs suggests that bureaucratic inertia stems from the fact "that established processes represent an enormous previous investment in time, effort and money" (Inside Bureaucracy , p. 195).

[4] Others have noted the importance of democratic beliefs. Robert A. Dahl, for example, argues that "the extent to which structural changes can achieve their ostensible purposes is severely limited by the civic orientations of citizens and leaders" (Dahl, Dilemmas of Pluralist Democracy [New Haven: Yale University Press, 1982], p. 133). "In short, the outlook for democracy depends on the

commitment of bureaucrats and politicians to democratic principles," Joel D. Aberbach, Robert D. Putnam, and Bert A. Rockman contend (Bureaucrats and Politicians in Western Democracies [Cambridge MA: Harvard University Press, 1981], pp. 170–71). Robert Jervis more generally argues for the importance of bureaucratic perceptions as a determinant of bureaucratic behavior (Jervis, Perception and Misperception in International Politics [Princeton: Princeton University Press, 1976]).

standing of bureaucratic attitudes toward the issue of democratic control therefore requires both looking inward at the agency and outward to society.

My goal in this chapter is to explain how bureaucrats reconcile the conflict between their two roles.^[5] I begin by demonstrating that the bureaucrats I spoke with do not reject traditional conceptions of democracy, although they seek isolation from the political process. I then proceed to examine how bureaucrats are able to hold such seemingly contradictory beliefs.^[6] I am not trying to assess the validity of the positions bureaucrats take. Rather, I am trying to understand in detail what their perspectives are. If these perspectives were as varied as the individuals involved, they might be of interest, but would be of little use. This is not the case. There are patterns to bureaucatic beliefs, patterns that are attributable to the daily experiences of bureaucrats on the job.^[7] An understanding

[5] For a general discussion of the issue of how belief systems arise to resolve role conflicts and then have an impact on how those conflicts continue to be resolved, see Peter L. Berger and Thomas Luckmann, The Social Construction of Reality (Garden City NY: Doubleday, Anchor Books, 1967).

[6] Robert D. Putnam's findings on the normative beliefs of members of Parliament in Great Britain and Italy support the expectation that there should be a relationship between attitudes about democracy and attitudes about control. He found that "the relationship between models of democracy and attitudes toward popular participation … shows a neat and unbroken progression as one moves along the continuum from authoritarian democrats to polyarchal and liberal democrats to classical democrats. Respondents at each step along this continuum advocate a higher and higher level of involvement and influence for members of the mass public" (Putnam, The Beliefs of Politicians [New Haven:Yale University Press, 1973], p. 203).

[7] Michael Lipsky makes a forceful argument in Street-Level Bureaucracy (New York: Russell Sage Foundation, 1980) for the importance of daily work routines in shaping bureaucratic practice and belief.

of these patterns in the way bureaucrats think about the issue of democratic control will reveal both paths toward improving the prospects for such control and pitfalls that might otherwise go unnoticed.

In order to look at the issue of democratic control through the eyes of the bureaucrat, I conducted a study of the attitudes of local bureaucrats. The subjects of my study were thirty-nine administrators who worked for the housing authority, fire department, or board of education in a moderate-sized northeastern city. In 1980 the city had a population of approximately 125,000, about one-third of whom were black and under 10 percent Hispanic; the median family income was about $15,000 a year. Almost 20 percent of the families had incomes below the poverty level.^[8] The city is governed by an elected mayor and by a city council selected in a partisan election on a district basis. During the interview period the majority of the council and the mayor were Democrats; a Republican subsequently assumed the mayoralty. Each of the agencies studied has a board appointed by the mayor that oversees agency affairs. Board members serve for fixed, staggered terms. The education and fire departments receive virtually all of their funds from the city budget, and the housing authority is funded largely from federal and, to a lesser extent, state funds.

I talked to all major central office decision makers, as well as to randomly selected chief field administrators (principals in education, deputy chiefs in fire, project directors in housing).^[9] ln education and housing the chief

[8] United States Department of Commerce, Bureau of the Census, 1980 Census of Population: Social and Economic Characteristics (Washington DC: Government Printing Office), vol. 8.

[9] In education I spoke with the superintendent, the assistant superintendents, the directors (of personnel, elementary schools, middle schools, secondary and continuing education, adult education, pupil services, special projects)

and several principals. In the fire department I spoke with the chief, the assistant chiefs, the deputy chiefs, the fire marshal, the deputy marshal, and the supervisors (of motor apparatus and buildings facilities). In housing I spoke with the executive director, the assistant directors, the directors (of maintenance, police, and projects for the elderly), the controller, and the managers of several specific programs (housing development, section, 8 tenant management).

administrator is hired by the board and can be dismissed only for cause during the course of the contract. The other respondents are career officials. All of the people I spoke with have considerable responsibility and authority. Most of the central office officials have administrative or technical staff working for them, and the field administrators have primary responsibility for delivering the agency's services. These are the people who design their agency's programs and oversee their implementation, who develop the agency's budget and negotiate with other government officials to obtain the necessary funds, who diagnose and solve problems in the agency as they arise. In other words, these are all people who make policy.

I chose to study local bureaucrats because in many ways they are the frontline troops in the battle for democratic control. The services they deliver may not be the most globally important, but they have an immediacy to people's daily lives that often leads citizens to care passionately about what an agency does, and to feel close enough to the agency to make their concerns known.^[10] During the period in which I conducted my interviews, each of these agencies was caught in a web of pressures from the mayor, council, community groups, client groups, and individual citizens over such issues as budget cuts, contracts for agency personnel, the adequacy of services in various neighborhoods, specific personnel decisions, school closings, class size, discipline, eviction procedures, lease provisions,

[10] See Douglas Yates, The Ungovernable City (Cambridge MA: MIT Press, 1977), for an elaboration of this argument.

and the quality of arson investigations. These pressures were amply reported in the local newspapers and reflected in the comments of the people I spoke with.

Interviewing thirty-nine people is clearly an exercise in hypothesis generation, not hypothesis testing. I selected respondents with an eye to obtaining a range of attitudes, not to creating a sample that could confidently be generalized to a larger universe. I chose the specific agencies I did because I thought they would provide variation on some of the dimensions I expected to be important in determining attitudes and in conditioning responses to control efforts by various political actors. The people I spoke with differed considerably in terms of such things as education, professionalization, the technical complexity of their jobs, and community support for the work of their agency.^[11]

Several factors suggest, however, that the patterns of belief I uncovered were not atypical. Despite the fact that my respondents varied in age, education, and background both within and across agencies, the similarity of their beliefs on many issues is striking. Furthermore, my interviews were conducted during a time of political transition in the city involved. As a result, in spite of their working in the same city, the political connections and loyalties of these bureaucrats were far from homogeneous. They were a mixed group whom it would be hard to call typical of the bureaucrats in a certain kind of city. Some had risen through the ranks; others had come into the agency laterally. Some had ties to an old-style ethnic political organization, some to newer "reform" leaders, and others virtually no local political ties at all.

The interviews themselves consisted of a series of

[11] A full discussion of the nature of variation among these agencies and the consequences of it is the focus of chapter 6.

open-ended questions^[12] about the individual's background, attitudes toward his or her job, the nature of his or her work, patterns of influence over agency affairs, attitudes about the proper role of various political actors in local decision making, attitudes about the general problem of democratic control of bureaucracy, and about the individual's conception of democracy.^[13] The interviews ranged in length from an hour to two and a half hours. Everyone contacted granted me an interview.

Overall, these interviews paint a challenging picture for those concerned with democratic control of bureaucracy. The bureaucrats not only describe themselves as being relatively free from such control, they also do not perceive much of a need for it. In their world the role of bureaucrats as political actors pales in comparison with the role of bureaucrats as people on the job. They see the strictures of democracy as being applicable to others, but not to themselves. They do, however, believe in those strictures; they espouse fairly classic democratic norms. But the bureaucrats manage to construct a view of their own situation such that, even when democratic norms seem applicable, they do not prescribe significant constraints on bureaucratic behavior.

Autonomous Bureaucrats, Democratic Citizens

The bureaucrats I spoke with consider themselves to have significant latitude of action, and they like it that way. The

[12] For a discussion of the use of open-ended interviews for this kind of research, see Joel D. Aberbach, James D. Chesney, and Bert A. Rockman, "Exploring Elite Political Attitudes: Some Methodological Lessons," Political Methodology 2 (1975): 1–28.

[13] A copy of the interview schedule appears as Appendix I. All but three of the interviews were tape recorded. All were then coded from transcripts either of the tapes or of notes taken during and immediately after the interviews. Coding was checked by a second person, who read a random sample of the transcripts with no appreciably different result.

weight of external control that they feel either through orders or, more subtly, through influence is slight. Thirty of the thirty-nine report that they "feel pretty much free to do what they want to do on the job."^[14] A school administrator explained, "I personally feel that I do have a great deal of freedom on the job in terms of making decisions about children, in terms of running the operation here." And from a housing administrator I heard, "I have the ability and the authority to operate the department as I see fit, as long as it conforms to the general goals and objectives of the authority itself." When asked who else had influence over department policy, only seven of the administrators said the mayor; fewer still said clients, the public, or other politicians.

These perceptions of freedom do not mean that the bureaucrats are totally unconstrained by outside forces. Consciously or not, the bureaucrats surely realize that there are certain types of behavior that would invite mayoral, council, or citizen response. Yet on a day-to-day basis these bureaucrats are clearly acting within the weak constraints such limits create.

Whether the resulting discretion is seen as excessive depends on one's view of democracy (as discussed in chapter 2) and on the desires of the other political actors involved. If this form of weak constraint were all that the mayor, the council, or the citizens ever wanted to impose, the bureaucrats would be under firm democratic control. On the other hand, if these actors sought to tighten or change the constraint, the issue would become whether the bureaucrats would accept this. To use Chester Barnard's famous phrase, the question is whether the bureaucrats are working within the citizenry's "zone of indifference" or whether the citizens are operating within

[14] All the data reported in this chapter are for the sample as a whole. Policy area effects will be discussed in the next chapter. Some quotations have been slightly changed to disguise the identity of the respondent or the city.

that of the bureaucrats.^[15] The answer to that question is suggested by the bureaucrats' description of how they think influence ought to be distributed.

Most of the bureaucrats I spoke with prefer outside actors to have very little power. When asked, "Who do you think has a right to a say about what goes on in your department?" the most commonly cited outside groups were those affected by decisions (clients, parents, tenants); but even such groups were mentioned by less than a third of the respondents. The community was included by ten, the board by six, and the mayor by only five. These figures probably indicate even greater unwillingness to allow outside influence than they at first suggest, since during the course of most of the interviews it became clear that the bureaucrats interpreted "say" very loosely as "complain" or "make their ideas known," and not necessarily as "constrain agency activities."

Each of the bureaucrats was also asked to describe the kind of role he or she thought the board, the mayor, and the public should play in the operation of the department. For the board, the bureaucrats sought the reverse of what is formally prescribed. Twenty-eight of the thirty-nine thought the board should merely react to agency proposals. A fire administrator told me his ideal role for the board would be one of "helping the chief efficiently run the department," and a housing administrator said of the board, "They can help in getting funds. They can be supportive when the problems come in." A school administrator was more expansive in describing his preferred role for the board: "I'd prefer to see the board show its confidence in the superintendent and his advisory staff, or in his total staff, and be inclined to take action on proposals

[15] Chester I. Barnard, The Functions of the Executive (Cambridge MA: Harvard University Press, 1938), chap. 12.

that have been professionally developed, [that have had] professional imprimatur, if you will, put on them and then [been] brought to the board for their consideration, review and attention." For the overwhelming majority of the administrators, the ideal is a board that basically serves the department, not the reverse.

Mayoral influence is even less welcome. Only five of those interviewed advocated active mayoral involvement in departmental affairs. The rest were fairly evenly split between those wanting the mayor merely to cooperate (seventeen) and those wanting no mayoral involvement at all (sixteen).^[16] The former view was typified by a housing administrator who said he thought the mayor and council members "should play a supportive role" and by the school administrator who told me, "I think there should be a cooperative kind of relationship so that we can understand what their problems are, carry out our responsibilities and also listen to what they feel is necessary." The latter, more closed view is reflected in the comments of an educator who said, "My feeling is that there shouldn't be any political influence on the schools," and of a housing administrator who told me, "Practically, I would say that it's easier to do a good job if the local political control is not there."

There was only slightly greater support for a role for the community, with about a third of the administrators reporting that some input short of decision making was appropriate. An educator described his preferred role for the public this way: "I think there should be input, appropriate input. Not necessarily on decision-making levels, but certainly on the collection of information and opinion." A fire administrator talked of a similarly limited

[16] Figures do not always total thirty-nine because some respondents did not answer all questions.

role for the public: "Oh, the public. Well, you certainly don't deny them having their input into it…. We would certainly honor any request that any citizen has. We would listen to them, try to answer their question." Two-thirds of the administrators advocated still more limited roles for the public—that is, either "being supportive" or no role at all. A housing administrator explained the former position: "Well, I think the general public has a dual role, that of being supportive of public housing … and also the role of informing one another of what public housing is and the fact that it is a type of housing that cannot be excluded from certain neighborhoods." A fire administrator typifies those who advocated no role for the public at all. He told me, "You just can't take people off the street and say 'Okay, we're going to give you a hand in running the department.'" Thus, whether the subject is the mayor, the board, the city council, or the public, these interviews strongly portray a group of bureaucrats who see power over agency affairs confined within the agency, and who believe that this is the way it should be, a happy situation for a bureaucrat, perhaps, but not for those concerned with democratic decision making.

A very different picture emerges, however, when the topic shifts from a discussion of an individual agency to beliefs about democracy. In spite of their distaste for outside intervention, these bureaucrats do not reject traditional conceptions of democracy. When I asked them about what they see as the essentials of a democracy, they raised the classic themes of self-governance and freedom. A fire administrator put it simply: "Government by the people, for the people. That's what it means." The sentiment was echoed by a school administrator. "You tell the politicians what to do. You elect your own officials. You vote them out when you want to vote them out." A colleague took this idea even further. "The Declaration of

Indepedence. It's an organization that provides government in which the consent of the governed plays a very, very strong role. And that's in all areas, not only in taxation or legislation."

Others stressed liberty. A housing administrator said of democracy, "I think I go along with the Constitution on that. The right to the pursuit of happiness, free speech, those kinds of things." A school administrator held democracy to mean that "I'm free to worship where I wish and free to vote as I wish." Often the themes were joined, as they were by a housing administrator. "I suppose equal protection. That a person be accorded the same rights or privileges no matter who he is, how much money he has, or doesn't have, where he lives. A democracy really has to protect the rights of the few while doing what the majority feels right."

These are hardly the words of people who question traditional democratic teachings. Only nine of the bureaucrats I spoke with raised the idea that a democracy required a government with a role independent of its citizens, and in every case this was qualified by the statement that government should be responsive. Furthermore, when asked, "Are there ways that this city or country could or should be more democratic than it is today?" only one person suggested that less popular participation would be desirable. In fact, the most common reform suggested (by eight people) was that popular participation be increased. Typical of such suggestions was that from a school administrator who told me, "You have to listen more. More input from the people." Others suggested electoral reform, greater socioeconomic equality, reducing corruption, and reducing corporate power.^[17]

[17] Similar patterns of belief have been found by other researchers. Joel Aberbach and Bert Rockman found that a majority of American federal executives were sympathetic to increasing the role of citizen activities, although

few of them had no reservations about the proposition ("Administrators' Beliefs About the Role of the Public: The Case of American Federal Executives," Western Political Quarterly 31 [December 1978]: 510). Aberbach, Putnam, and Rockman report on the beliefs about the role of the public held by European bureaucrats. They found that 44 percent of them were either rather or very favorable to increasing popular control or citizen participation and that only 16 percent of them thought that the role of the public should be confined to voting or less (Bureaucrats and Politicians , pp. 182–83).

Support for the basic tenets of democracy does not, however, necessarily produce an appreciation of how those tenets are translated into the actual workings of government. In fact, the bureaucrats I spoke with seemed almost blissfully unaware that their desire for autonomy might conflict with democratic values. Fairly late in the interviews I asked, "Some people talk about public agencies being undemocratic. Do you think that this is a problem?" The response was strongly, "No." Fewer than a third said they thought such a problem existed. Since further conversation revealed that many of these were not talking about control at all, this probably overstates the number who saw the contradiction between democracy and bureaucratic decision making. A school administrator, for example, said that he thought there was a problem with public agencies being undemocratic, and explained, "That means you're not sensitive to people's needs." His proposed solution to the problem was simple: "Someone calls and you answer the telephone. You talk to them." Similarly, a housing administrator proposed "proper communication, notification, considering your fellow man" as ways to make public agencies more democratic. And a fire administrator, who thought his agency was in fact very democratic, offered this example: "We're very democratic around here. We try very hard to be part of the public. We, I think, sometimes even bend over backwards. For example, the guys take up donations to help fire victims." Answering phone calls and helping fire

victims may be desirable, but the fact that these were the primary examples of the democratic nature of public agencies these bureaucrats could offer testifies to the weakness of their conception of the ways democracy interacts with their work.

Approaching the question of democratic control from a different direction, I asked the bureaucrats whether they felt public agencies were too responsive to outside pressures and groups. The answers give further evidence that on an intellectual level the bureaucrats do not see the need for greater democratic control. Almost half of the administrators felt that excessive responsiveness was not a problem, but only because agencies could hold their own against outside pressures. As a fire administrator flatly said, "We don't get pushed around." A housing administrator was gentler, but made a similar point: "Responding, but not necessarily giving in, is what I see going on here, and I think that's healthy." Only two bureaucrats said they felt public agencies were not responsive enough. One-third replied that public agencies were already too responsive. A fire administrator stated simply, "I think that is a problem. I think they've gone a little too far and too fast, which is not good." A school administrator answered much more personally: "There's no question in my mind we react too much to community pressures. And I think that many decisions that we make are made almost on the spur of the moment in terms of whatever the pressure from the community is." Another school administrator made a similar point: "We have some principals who are so concerned about being open and responsive to everybody that they are really nothing to anybody. They're bounced from pillar to post. They are appeasing this group and that group and the other group. There's no common denominator. There's no pattern. There's no strand, no cohesiveness about any of their actions."

Still a third tack was to raise the specific problem of red tape, a most galling issue to the public, and one that in its saliency and simplicity might be expected to be particularly evocative of responses compatible with democratic concerns. The bureaucrats were not too sympathetic. Only ten said without major qualification that red tape was a problem; four allowed that it might be a problem elsewhere, but not in their own agency; eight said it was not a problem at all; and twelve argued that the rules embodied in red tape were necessary. Characteristic of this last perspective was the housing administrator who explained, "Most of the forms we have to fill out and the procedures we have to go through I think are pretty necessary … to ensure that the housing authority runs properly, to ensure that all people are treated fairly, and for our own records, for our own knowledge of what is happening." A fire administrator made a similar argument about red tape: "There are certain ground rules that everyone has to follow … sometimes it's important." Thus, regardless of whether the issue of democratic control was raised directly or negatively (and in a weak form—i.e., responsiveness—at that) or by a specific problem, the response of these bureaucrats, who otherwise subscribed to democratic beliefs, seemed to be, "Who, me?"

How can bureaucrats accept democratic principles yet reject their implications? Why is it that bureaucrats do not seem to see a conflict between their normative beliefs and the way they seek to live their daily lives? The answer lies not in a problem with the beliefs, but in the other half of the question, in the daily lives of the bureaucrats. Bureaucrats are not only appointed government officials; they are also people at work. Bureaucratic perspectives on democratic control seem to be affected much more by the ongoing realities of the job than by abstract beliefs.

Such beliefs may be of central importance to the would-be controller's choice of a control strategy, but they are largely irrelevant to the bureaucrat's response.

The conditions bureaucrats work in and their perceptions of those conditions do two things. First, they tend not to raise democratic norms, and thus fail to provide an environment in which these norms can guide behavior. Second, they redefine the first two components of the concept "democratic control of bureaucracy"—that is, democracy and control—in ways that create variances for the specific situation of the individual bureaucrat. These variances then enable bureaucrats to make their opposition to control consistent with democratic beliefs.

The Insulated Bureaucrat

Bureaucrats are public officials, but they are public officials who do not go through the socializing ritual of the electoral process. They have neither to proclaim American democratic verities in order to keep their jobs nor go through an overtly political process to obtain the jobs in the first place. Thus it should not be surprising if bureaucrats rarely think of themselves as political actors subject to democratic strictures. In fact, the bureaucrats I talked with rarely spoke of themselves as engaged in the political process and almost uniformly failed to connect their work with the operation of a democracy.

"Tell me a little bit about how you see your job," I asked my respondents. Their answers were predictably varied, but two-thirds described themselves in terms that added up to their being administrators. Some listed their specific responsibilities, such as budget preparation personnel management, maintaining records. Others gave a more general response, like the school administrator who said,

"I have major responsibility for the administration and supervision of a portion of the schools in this city…. I think one of the main roles I'm involved in is to ensure that the schools function in an orderly, meaningful educational manner." Two emphasized that they provided a substantive service, including a school administrator who replied, "I see my job as more or less giving someone an opportunity to do or to be what they want to do." A fire administrator typified a final group of eleven by giving an affective description: "Well, I enjoy it. I find it exciting and challenging."

What is most interesting about these responses is that no one described him or herself as a public official.^[18] Only eleven of the bureaucrats even mentioned the public when they described their jobs, and only two said anything suggesting that policy was a part of what they did. In fact, during the course of the interview, three-quarters of the bureaucrats at some point asserted that they did not make policy, that policy was made elsewhere. Overall, then, these bureaucrats do not give the impression of a group aware of their position in a democratic political process. In the words of one of my respondents: "I keep away from politics. Politics are not my cup of tea."

As they see it, the world of bureaucrats is filled with administrative matters, technical services, substantive programs and job satisfactions or dissatisfactions—not with political issues or the search to implement democratic norms. Robert Putnam discusses a related phenomenon in his work on the attitudes of European civil servants. He argues that civil servants "use technical or

[18] Robert A. Dahl and Charles E. Lindblom argue that "faulty identification" puts limits on the external control of bureaucracy. The identifications they discuss are as a professional and as a member of an organization (Dahl and Lindblom, Politics, Economics and Welfare [Chicago: University of Chicago Press, 1976], p. 260).

administrative or financial criteria for defending their preferred policies and for criticizing alternatives…. Politicians live in a world of political pressures to be reconciled, while civil servants live in a world of practical problems to be solved."^[19]

A self-image of technician or administrator can, of course, be useful for preserving one's autonomy vis-à-vis political leaders. If what one does is not political, then politicians have no business intervening."^[20] The obvious expedience of this posture for bureaucrats invites scrutiny and, perhaps, disbelief. But I do not think that this apolitical persona is contrived, a ruse elaborately designed to protect bureaucratic turf. My respondents initially had no idea I was interested in the problem of democratic control. I approached them saying I was doing a study about bureaucratic life and I wanted to interview them about their jobs, phrases purposely meant to be vague. The question asking them to describe their job was one of the first asked, long before any mention was made of mayors, citizens, or democracy. Thus, even if the myth of the apolitical bureaucrat had begun as a ruse, it was one that had become so incorporated into the thought processes of virtually everyone I spoke with that it was expressed in response to the most innocent question. Whether born as a widespread dissimulation or, as I believe

[19] Robert D. Putnam, "Bureaucrats and Politicians: Contending Elites in the Policy Process," in Perspectives on Public Policy Making , ed. William B. Gwyn and George C. Edwards Ill, Tulane Studies in Political Science, vol. 15 (New Orleans: Tulane University Press, 1975), pp. 191, 202. Aberbach, Putnam, and Rockman advance a similar argument (Bureaucrats and Politicians , chap. 5).

[20] Ezra N. Suleiman discusses this in relation to the highest level French administrators. He argues that "the belief that the politician ought not to enter into the details of the work of the Directors reflects a significant self-definition of the civil servant's role: a technician with a monopoly over a domain. A politician either does not infringe upon this monopoly because of his other, more enduring interests, or he cannot because of lack of competence" (Suleiman, Politics, Power and Bureaucracy in France [Princeton: Princeton University Press, 1974], p. 231).

more likely, in response to the normal pressures and environment of bureaucratic life, the myth has become genuine belief.

A glimpse at the daily life of these administrators reveals why they think of themselves in this way; there is little to remind them that they are, in fact, public actors. They rarely communicate with political leaders, and neither do they have much direct contact with them. These bureaucrats thus do not have a substantial opportunity to connect their work with broad questions of politics, much less with democracy.^[21] I asked a series of questions about the individual's patterns of communication and interaction, and the answers consistently pointed to insularity within each agency. When asked to whom they talked about their work, the most common response (by twenty-six of the bureaucrats) was other people in the agency. The next most common answer was a relative or friend (given by sixteen of the thirty-nine). In contrast, only two of the bureaucrats mentioned the board; three either the mayor, someone from his office, or a member of the city council; two the community; and three clients. Similarly, when asked whose advice they sought, the answer was rarely elected officials. Rather, it was most commonly others in the agency (by twenty-one) and sometimes fellow professionals (by eight). Only two included the board in their list of those they asked for advice, two included clients, and no one mentioned the mayor or any other political figure. In neither of these questions were the bureaucrats

[21] Aberbach, Putnam, and Rockman discuss the relationship between bureaucratic contacts and political attitudes from the other direction, that of the bureaucrat whose attitudes are unusually close to those of politicians. They found that support for pluralism and for populism is unusually high among those closest to the center of power: "It is as if contacts upward and outward from the bureaucracy remove them from a climate that is less hospitable to democratic ideals and expose them to an atmosphere more like that ordinarily breathed by politicians" (Bureaucrats and Politicians , pp. 203–4).

forced to make a single choice; they were allowed as many responses as they wanted. Thus the low numbers for contact outside the agency do not reflect those for whom this is the primary source of interaction, but those for whom there is any such reported interaction at all.

The pattern is reinforced from outside as well. I asked the bureaucrats, "Who asks you for advice?" and the answers were virtually the same. Overwhelmingly, the people who ask bureaucrats for advice are other bureaucrats in the same agency. Only one reported that board members asked him for advice; two said the mayor or a member of the city council did; two said citizens. Perhaps not surprisingly, those mentioning a board member, the mayor, or some other public official were almost exclusively agency heads. This is probably the result of a combination of administrative courtesy and good sense that dictates to outsiders that the agency chief not be bypassed. An unfortunate side effect, however, is that a context for allowing administrators in general to feel that they are part of the political system is forfeited because of the inability or unwillingness of political representatives to penetrate the bureaucratic world.

Bureaucratic isolation is also revealed in the bureaucrats' reports of the frequency of their contact with board members, the mayor's office, and members of the city council. Almost half reported that they seldom had contact with the board. Those who did have contact described it as being dominated by requests for specific information and not by discussions of programs or policies. Only three said they ever discussed policy questions with members of the board. In order to explain the paucity of their interactions with the outside, administrators of both the housing authority and the fire department stressed that lines of authority ran through the head of the agency.

But none of the three who reported having policy discussions with board members was an agency head.

Lines of communication with the elected officials of the city are also weak. Six of the bureaucrats reported fairly frequent contact with the mayor's office, but this was virtually always about complaints the mayor had received or about requests for information. A school administrator described his contact with the mayor's office as being "just problems," and a housing administrator told me that her only contact came from the fact that "They write the director all these incredible memos about complaints, and I usually get to answer them." Only six of the administrators reported ever having contact with the mayor about programs or policies.

About half of the bureaucrats reported some interaction with members of the city council, but again the interaction was overwhelmingly about constituent complaints and problems. A housing administrator described his contact with the council as "periodic. And it's mostly when there's trouble." A fire administrator also reported periodic contact: "Some councilmen, because they represent their constituents … will say 'Take a look at the building for me. There are some people in my district who are interested in it. If you do, I'll appreciate it.'" A colleague in the fire department reported "no contact, not unless I get calls. At times we have heavy rain storms, or heavy floods, or a condition that a councilman may be trying to correct." Only two of the administrators indicated that they had regular or frequent contact with a member of the council about policy matters. Those lines of communication that do exist primarily concern adjudication or amelioration of specific problems, not policy issues. Such patterns of contact reinforce the bureaucrats' conception of themselves as implementers, not policy

makers, since discussions are over narrow details, not broader trends.

The Constrained Bureaucrat

Daily life on the job may do little to remind bureaucrats that they are public officials who should be democratically controlled, but it also does little to suggest to them that lack of control is a problem. They feel constrained, but not by people outside their agencies.^[22] When asked, "Who sets limits on you?" the large majority indicated that limits were derived from within the organization. As a fire administrator told me, "I have the chief to answer to. Anything that comes up as far as a major decision naturally has to be taken to him." A colleague in education described a similar situation: "I do make it a habit of checking with the superintendent, but essentially I can act, and not react." Limits on housing administrators were likewise internal to the agency: "I would say that on any substantive issue there really aren't limits as long as the director knows what I'm doing. Those are more or less the only limits." Only one bureaucrat felt sufficiently unconstrained to say that the primary limits were self-imposed, but only a handful replied that they felt limits from outside the agency—five replied that the board imposed limits and one that the mayor did. Each of the bureaucrats was also asked, "Who do you see yourself as being answerable to?" Only six gave the highly autonomous reply that they felt responsible merely to themselves. Thirty of the thirty-nine reported that they were responsible to

[22] See Herbert Kaufman, Red Tape (Washington DC: Brookings Institution, 1977) for a description of many of the constraints bureaucrats face.

their superiors in the department. They were allowed to list as many people as they wanted, but still only four included the mayor, the city council, or any other political leader, and only thirteen included the board.

If these bureaucrats worked in agencies that approached a classically ordered hierarchical organization in which all power flowed from the top to the bottom, these results might not be troubling from a democratic perspective. The heads of the agencies would be the ones subject to outside influence, and their subordinates would feel that influence only as constraint from within the agency. The control would be substantial enough, moreover, that lower level bureaucrats would exercise little discretion. In fact, however, the agency heads seemed to feel only slightly more constrained by outside forces than the other bureaucrats and, while none of the bureaucrats felt free of constraint, they all reported exercising a fair amount of discretion. Of the three agencies I studied, the fire department came closest to the hierarchical model. Yet the fire chief told me, "I have complete freedom. I couldn't operate any other way. There is a board of fire commissioners … but they don't make the decisions about the fire department. They discuss things with me but they don't make the judgments." The superintendent of schools reported feeling the influence of the incumbent mayor, but saw him as an aberration. Of the previous administration he reported, "It was such a pleasure to run the school system. When you don't have interference, it's beautiful, it's just great." The external constraint felt by the head of the housing authority came not from citizens or elected officials, but from other bureaucrats—those in the federal government.^[23]

[23] Further doubt is cast on the hypothesis that democratic control flows through the top of an agency by the fact that none of the six people who reported contact with the mayor's office over policy issues is an agency head.

Furthermore, there is no obvious hierarchical logic that would predict who those six are. Overall, the agency heads do not stand out in terms of their patterns of interaction and communication.

Redefining Norms

The world in which bureaucrats work may do little to suggest that democratic norms should govern bureaucratic behavior, but bureaucrats are not wholly oblivious to their governmental role. When they do think of it, however, they manage, through a variety of what might be called "sleights of thought," to redefine two key concepts—control and democracy—so that bureaucratic behavior conforms with a newly constituted norm of democratic control of bureaucracy.

Control is dealt with quite simply: it is conceived of in such a manner that its exercise by an outsider does little to constrain what bureaucrats do. One way bureaucrats accomplish this is by accepting the distinction between policy and administration but then defining policy in such a way as to make it an almost empty concept. As noted earlier, the bureaucrats I interviewed overwhelmingly rejected the idea that they made policy. Policy according to them is made elsewhere, usually by the board. If this were true, then democratic control of bureaucracy would be less of a problem. Elected officials or their surrogates on boards and commissions would make policy, and bureaucrats would merely do what flowed from that policy. Unfortunately, as virtually any student of administration knows, it is not so simple to distinguish policy from administration, and neither is it automatic that making "policy" means significantly controlling administrative discretion. This is particularly true when administrators themselves have a large role, as they often do, in determining what falls within the realm of policy, what within that of administration. "Most issues are defined by superintendents

as internal, and many boards appear to have difficulty in staking out a legitimate territory," argue Harmon Zeigler and M. Kent Jennings, writing about local educational decision making.^[24]

It should come as little surprise to learn that to the bureaucrats I spoke with, policy making is neither a very strenuous activity nor one likely to result in major constraint on administrative behavior. In their eyes, policy making consists of specifying the most general outcomes the agency is to achieve; all the rest is administration. An educational administrator described it nicely in response to a question about the role he saw the board of education playing: "I think that making education policy is their major role. I think the way the system should operate is that they do just that but limit themselves to just that. They make policies and from there on in it becomes administrative." I then asked him for an example of a policy decided by the board. His answer: "Teach kids to read."^[25]

A second variation on this theme is that policy making may involve making decisions that impose greater constraint on bureaucratic behavior, but that these are purely reactions to proposals from the administrators themselves. A housing administrator explained: "The board should ratify policies but take no operational role." Another educational administrator was somewhat more expansive. When asked to give an example of policy making

[24] Harmon Zeigler and M. Kent Jennings, Governing American Schools: Political Interaction in Local School Districts (North Scituate MA: Duxbury Press, 1974), p. 157. See also Frederick M. Wirt and Michael W. Kirst, Political and Social Foundations of Education (Berkeley CA: McCutchan, 1975), chap. 5. In Politics, Planning and the Public Interest (New York: Free Press, 1955), Martin Meyerson and Edward C. Banfield discuss at some length the ways the Chicago Housing Authority staff influenced policy making by commissioners.

[25] Similarly, Zeigler and Jennings argue that superintendents define boards' choices as being either support or firing. Any supervision of the educational program is defined as "being involved in administration" (Governing American Schools , p. 190).

by the board, he suggested a situation in which a superintendent met with high school principals and counselors and concluded that the curriculum should be more demanding. The superintendent would then

propose and recommend to the board that they set a policy that no youngster will graduate from the high school unless he has twenty units instead of sixteen. That's policy. He presents it, they review it, they interrogate him, question him why he wants to do it—back and forth—are there any problems? No? All right, no problems. Then somebody makes a motion "I move we accept this policy and put it in our administrators' manuals." And there's a second, they vote, now it becomes policy.

It may comfort administrators to think that they are not making policy and are therefore immune to the problems of democratic control. But observers who value democratic control will not be appeased. In effect, by granting a board such a limited policy-making role, the administrators are weakening the board's authority, while absolving themselves of charges that they are straying into the political realm, where they agree they do not belong. They thus accept extremely weak externally imposed constraint and reject all other political control as illegitimate.

Administrators have a second way of accepting democratic control by weakening its meaning: they transform control into responsiveness to individuals. The bureaucrats I spoke with had fairly extensive contact with citizens and expressed fairly strong acceptance of the need for it. Over half said that they had either daily or frequent contact with individual citizens or citizen groups, and two-thirds felt that this was an important part of their job. Less than a quarter said that they rarely or never had such contact, and only two suggested that these contacts were a necessary evil. But the nature of the contact was highly particularistic; it did not concern policy issues. Fire administrators

described telephone calls about air conditioners, bats, fireplaces, sewers, and the fire code. School administrators talked with parents about specific problems their children were having and with citizens who lived near schools. Housing administrators fielded tenants' questions about maintenance, evictions, and damages. By being open to this kind of interaction with citizens and accepting the need for it, administrators are able to believe they are fulfilling their responsibilities as governmental officials while simultaneously denying external actors significant control over their behavior.

Equally revealing are the administrators' answers to questions about why they think contact with citizens and citizen groups is important. Over and over, the emphasis was on helping individuals, on being humane. One school administrator explained, "I think it is very important. If people take the time to call me or to write to me, they're in need, they want something, so I feel I should take the time to answer them." A housing administrator maintained, "I really resent people who put individuals off. I think that it's important that I develop that kind of rapport on an individual basis with people." And from a fire administrator, "I don't mind it, don't mind it one bit. I think if they take the time to call up, they feel something is wrong, and I see nothing wrong with them complaining."

Occasionally someone suggested that such responsiveness was particularly important, since citizens were also taxpayers, but even these comments reflected a conception of control as responsiveness to individuals, not an appreciation of the legitimacy of tighter externally imposed constraints. None of these people, nor any of their colleagues, suggested that his or her behavior was significantly changed by interactions with citizens; mostly they just gave answers or heard people out. But because they

did give answers and did hear people out, these bureaucrats did not think of themselves as isolated, untouchable figures. The problem of democratic control, therefore, is at least partly defused by allowing for responsiveness to individuals.

If with one hand administrators strip the concept of control of much of its meaning, with the other they alter what is generally thought of as its democratic element. Far from accepting democratically elected officials as legitimate agents of such control, the administrators I spoke with saw themselves as upholding democracy against the particularistic incursions of these officials. In bureaucratic eyes, mayors, board members, and members of the city council are not elected leaders of the citizenry, but rather politicians. And politicians are surely not the people to uphold democracy.

Many scholars have found that bureaucrats often have little use for politicians. They see them as privatistic, concerned only with their own electoral welfare or the interests of a narrow constituency. They are judged to be shortsighted, to believe that ends justify means, and generally to compare unfavorably with the bureaucrats themselves, who apply the same standards to all citizens in pursuit of the public interest. Discussing the French administrative elite, Ezra Suleiman explains, "For the Directors, a minister is first and foremost a politician. This means that his interests are always segmental, whether they concern the interests of his party, his local constituency, or even, for that matter, his ministry. They are never the general interest. The Directors see themselves, on the other hand, as the guardians of the general interest."^[26] Robert Putnam writes of the "classical bureaucrat" who "distrusts or rejects the institutions of politics, such as parliaments,

[26] Suleiman, Politics, Power and Bureaucracy , p. 232.

parties, and pressure groups. To the classical bureaucrat, the noisy, incompetent, partisan practices of politicians seem at best senseless, at worst positively inimical to the permanent interests of the state."^[27]

The bureaucrats I spoke with forcefully joined their voices to those quoted by other scholars. One educational administrator made his views known quite succinctly: "Politicians I abhor. I've seen many of them give half truths in order to accomplish their causes and without any thought or care evidently to what happens in the end." A colleague concurred: "Political groups gain power based on very short-term popular issues. Educational decisions affect a long period of time. It's much easier for the local political structure to do something that people want at a given time because it makes them popular for the next election." And still another colleague, discussing the mayor, maintained that "the mayor should be the farthest away from education. While he ought to be a member of the board, I don't think he should have a vote. He should be apprised on a monthly basis. That would give him the feedback and at the same time keep him away from education…. Once the mayor gets involved in education, then it becomes a political thing. And once it becomes a political thing, there's no end to it."

Lest one think that there is something peculiar about education operating here, listen to a housing administrator talk about political leaders:

I've seen firsthand some of the incredibly bad things which have happened at this authority as a result of boards that are not interested in the authority but are only interested in serving

[27] Robert D. Putnam, "The Political Attitudes of Senior Civil Servants in Western Europe: A Preliminary Report," British Journal of Political Science 3 (July 1973): 259. Aberbach, Putnam, and Rockman similarly describe bureaucrats who "are dismayed by what they see as the chaotic collision of selfish interests, obscuring and subverting the commonweal" (Bureaucrats and Politicians , p. 149).

some alien political interest in the mayor's office…. Most boards, even so-called good boards, are normally going to be somewhat responsive to the appointing authority, and the appointing authority is political by nature. I suppose that has the potential to present problems, and if there was any way to restructure that, I would.

A colleague's comments suggest that from the bureaucrat's point of view, the problem lies not merely with local political leaders but with national ones as well. "Politics enters into things more often than it ought to…. Not that I think I'm an oracle of Delphi and can perceive everybody's best interest, but I can perceive it better than Congress can. And they're only interested in it from a political standpoint." And a fire administrator perhaps summed up the feelings of many when, discussing politicians, he concluded, "Thank God they leave us alone."

With such an opinion of political leaders, it is small wonder that democratic control by elected officials makes little sense to bureaucrats. Although no one said so in so many words, it was clear that to these people such control was tantamount to asking the fox to guard the chicken coop.^[28] If anything, they see their role as guarding the public weal against the designs of politicians, and not as achieving it through those designs. Bureaucrats deny elected officials the right of control and deny it on the grounds that they are unfit governors, incapable of discerning the public interest. That is a task better left to the

[28] Some might think that local officials, the heirs of Tammany Hall, are particularly likely to bring such perceptions on themselves. Federal officials, however, have proven themselves extremely capable of particularistic action as well. Morris P. Fiorina even entitles one section of his book on Congress "Tammany Hall Goes to Washington" (Fiorina, Congress—Keystone of the Washington Establishment [New Haven: Yale University Press, 1977]. Douglas Arnold also provides evidence of such particularism in congressional behavior (Arnold, Congress and the Bureaucracy [New Haven: Yale University Press, 1979], chaps. 6–8).

experts. Since it justifies a large measure of autonomy in bureaucratic eyes, this position can, of course, be self-serving. But political leaders also bring this on themselves. Every time a mayoral aide or a city council member calls a bureaucrat about a constituent problem, he or she reinforces the idea that the concerns of political leaders center on petty issues.^[29]

By thinking about control as a greatly attenuated policy-making function or as minimal bureaucratic receptivity to individual citizens' complaints, bureaucrats are able to make their own jobs compatible with their broad democratic beliefs. By rejecting elected officials as pursuers of the public interest, bureaucrats are able to justify their rejection of interference by such officials and still be consistent in their belief in democracy. But what of the citizens themselves, the other potential agents of democratic control? How do the bureaucrats align their reluctance to admit control by the citizenry with their belief in democracy? The answer lies in a final sleight of thought that creates an exemption from democratic strictures for the specific agency involved. This enables the bureaucrats to accept democratic norms but deny their applicability to their own circumstances.

As I noted earlier, bureaucrats are somewhat more willing to grant the right of access (if not control) to citizens than to elected officials. Unfortunately, the citizens bureaucrats actually run into are often disappointing; they do little to suggest that they are truly capable of governing.^[30]

[29] Suleiman reports a similar concentration on particularistic dealings in deputy-director interactions in France, with similar results for bureaucratic attitudes. He found that 75 percent of the civil servants questioned believed that the deputy "was not concerned with questions of general policy but only with making demands on behalf of his constituents" (Politics, Power and Bureaucracy , p. 291).

This sentiment was captured by an administrator in the housing authority who complained that "the reality that I have become accustomed with is [that] people who might otherwise have valid opinions, don't." Talking about the role the public should play in his agency, a fire administrator explained, "We're answerable to everybody if they have a legitimate reason. Mostly though it's personal gripes, something for personal gain. They don't come in and talk about their fellow citizens." And, when it is not personal gripes, bureaucrats often associate citizens' demands with the controversial issues of the day. At the time I was interviewing, affirmative action in hiring was such an issue, and the word "democracy" would recurrently set a respondent off on what were often tirades against affirmative action. Identified as the stalking horse for special interests, democracy itself thus loses its strength.

As a result of this dim view of citizens' behavior, many bureaucrats conclude that at times they must safeguard the public interest not only against the demands of politicians but also against parts of the public itself because of flaws in the public's ability to discern what is best. Talking about community groups, a school administrator told me that "because of pressure you have to respond in a different way [from the way you think is best] and possibly the way the public feels. And of course you do that but sacrifice the whole for the part." A colleague was less willing to make the sacrifice. In discussing the role of the public, he argued, "You listen to a problem, but when it comes time to make a decision, [you] make it on a hell of

[30] Lipsky argues that street-level bureaucrats "mentally discount their clientele so as to reduce the tension resulting from their inability to deal with citizens according to ideal service models" (Lipsky, Street-Level Bureaucracy , pp. 140–41).

a lot more factors than what is politically popular at the moment."

In bureaucrats' eyes, popular ability to govern is also tempered by structural barriers posed by the complexity of their specific policy areas. Even if citizens were not be-having in a privatistic fashion, it would be difficult for them to comprehend fire, education, or housing policy. Three-quarters of my respondents maintained that the average citizen did not understand the work of their agencies; only four unreservedly thought the average citizen did.^[31]

Assertions and data such as these might lead one to question whether bureaucrats really believe that people should govern themselves at all. Yet in spite of this skepticism about the motives of the public, these bureaucrats had a striking faith in people. "In general, do you think people know what is best for themselves?" I asked them, and only four said no. Fifteen gave an unqualified yes, while the rest hedged on a basically affirmative answer. The problem lies not with the general belief, but with its applicability to the particular circumstances confronting the specific bureaucrat. When asked about their own policy area (education, fire, or housing) the pattern of bureaucratic opinion about citizens' ability to make decisions was a mirror image of that for the general question. Thirteen said no, people did not know what was best for themselves; only three said yes; and the rest said, "Yes, but…." As one administrator told me with perhaps surprising candor, "Oftentimes I think bureaucrats start believing

[31] Others, too, have found that the reservations of bureaucrats about citizen knowledge and understanding are major impediments to increasing democratic control of bureaucracy. See, for example, Dale Mann, "Democratic Theory and Public Participation in Educational Policy Decision Making," in Frederick M. Wirt, ed., The Polity of the School (Lexington MA: Lexington Books, 1975); Aberbach and Rockman, "Administrators' Beliefs"; and Aberbach, Putnam, and Rockman, Bureaucrats and Politicians , chap. 6.

that they know better what is best for people than people know themselves."

Normative beliefs rest on certain assumptions—in this case that people are sufficiently expert, experienced, informed and farsighted to determine their own interests. Most of these bureaucrats seem to accept the validity of these assumptions in enough cases to accept the norm, but their experience in their own work leads them to reject it for their agencies. These data indicate, however, that such reservations are seen as exceptions and not the rule. Thus, the democratic norm of the competent citizen is maintained; it is just not relevant to the bureaucrat's particular case.

Our excursion into the terrain of bureaucratic beliefs has yielded useful information about the prospects for democratic control, with some potential openings on the road to reform. The path of moral exhortation seems notably doomed to failure. My conversations with bureaucrats have convinced me that appeals for cooperation based on democratic norms are likely to be met with nods of agreement when the norms are stated and acts of disagreement when it comes to daily behavior on the job. For that reason, too, emphasis on educating bureaucrats or hiring people with different values is also likely to produce frustration. The values these bureaucrats have are fine; they are just not heeded.^[32]

In many cases the bureaucratic explanations for why norms are not acted on are clearly self-serving. By impugning the motives of elected officials and of ordinary

[32] Lipsky argues that "popular wisdom often identifies the source of workers' attitudes toward clients and their jobs in prejudices acquired in upbringing and social background. Such perspectives lead to recommendations to hire better educated personnel or provide further education and training in public and human relations. All too often such perspectives fail to take account of the influence of street-level bureaucrats' work on their attitudes" (Lipsky, Street-Level Bureaucracy , p. 141).

citizens, the bureaucrats justify their own relative autonomy. Yet it would be a mistake to discount totally these explanations in thinking about reform. Self-serving or not, what bureaucrats believe shapes what bureaucrats do in response to control efforts.

Seeing the world through the bureaucrats' eyes also gives us a window through which to view how well the normative assumptions underlying various forms of control are, in fact, realized. In this way, bureaucratic attitudes mark perils on certain paths of reform. To the extent that bureaucrats are correct that the concerns of elected officials gravitate toward their own political gain, these beliefs signal a problem with increasing mayoral or council control over bureaucracy.^[33] Similarly, to the extent that the bureaucrats are right about citizens being equally shortsighted, control by citizens that assumes that citizens will promote the collective interest is likely to go awry.

Finally, an understanding of the ways bureaucrats' daily lives affect the way they perceive the issue of democratic control suggests new ways of thinking about achieving that control. If one difficulty in bringing about control lies in the nature of the interactions bureaucrats have with the outside world, then one avenue for change may lie in changing those interactions. But to understand fully how this might be done, we first need to explore the ways in which bureaucracies, and hence the daily lives of bureaucrats, vary.

[33] Heclo warns of the federal bureaucracy that "there are greater dangers in any reform that concentrates exclusively on responsiveness to political leadership. Without a sense of the civil services' independent responsibility to uphold legally constituted institutions and procedures, political control of the bureaucracy can easily go too far. Any single-minded commitment to executive energy is likely to evolve into arbitrary power" (Heclo, Government of Strangers , p. 244).