Bureaucracy

PREFACE TO THE NEW EDITION

ONE OF THE CENTRAL THEMES of this book is that the management of government agencies is powerfully constrained by limitations on the ability of managers to buy and sell products or hire and fire people on the basis of what best serves the efficiency or productivity of the organization. Laws and regulations limit how people can be hired, greatly reduce the chances of firing anyone, and surround the buying and selling of buildings and equipment with countless rules about fairness and procedure.

There have been efforts over the last decade to alter these constraints. President Bill Clinton and Vice President Al Gore sought at the start of their first term to convert federal agencies into a government that “works better and costs less” by reducing red tape, inducing bureaus to put customers (that is, citizens dealing with the government) first, and empowering employees to get results. This effort was officially known as the National Performance Review (NPR) and popularly known as Reinventing Government, or REGO.¹

The NPR effort marked a sharp departure from past reorganization plans. Of the seventeen presidents to take office between 1904 and 1992, eleven created reorganization task forces that were chiefly devoted to expanding accountability, increasing efficiency, and enlarging presidential authority.² (By accountability I mean increasing policy implementers’ responsiveness to higher officials who must take responsibility for that policy.) This effort had some successes—for example, it equipped the president with more staff assistants and made more rational the way in which budgets were drawn up—but it did little about red tape.

The NPR began with very different assumptions. Where earlier reforms had stressed the accountability of officials to higher authority, the NPR emphasized agency responsiveness to the public. Where earlier reports relied on appeals to business efficiency, the NPR drew inspiration from the concept of entrepreneurial culture. And where preceding efforts urged an expansion of the power of the president, the NPR called for empowering government employees and restoring public confidence in government.

The dominant theme of the NPR was to solve the “root problem” of modern government—its excessive reliance on “large, top-down, centralized bureaucracies.” This solution necessitated the creation of “entrepreneurial organizations” that have “constantly learned, innovated, and improved.”³ The NPR meant, in effect, marrying the youthful spirit of Silicon Valley to the old traditions of official Washington.

The main reason for this startling shift in federal reorganization efforts, however, was not admiration for clever young people who made computer chips or invented Internet sales programs but, rather, a desire to reverse the sharp drop in the public’s confidence in the federal government. In 1964, three-quarters of the American people said they trusted Washington officials to do the right thing; today, fewer than one-quarter feel that way. Perhaps, the NPR staff thought, these citizens would like officials better if they “put people first” and “cut red tape.” For that matter, government bureaucrats might like their own jobs better if they could put people first and cut red tape.

But putting people first is hard to do in a government that, ultimately, has the power to command people and even send them to prison. A business may put people first because businesses compete with each other in order to attract customers, but the government competes with nobody. And cutting red tape may be possible in a business firm that can tell whether it is doing a good job by looking at its sales and profits, but cutting it in a government agency is much harder because (ordinarily) government agencies deal with neither sales nor profits.

Changing government is difficult, but serious efforts have been made. Let me describe two important changes that should modify what you read in this book, which was written several years before the NPR had started.

Procurement

At several places in this book,⁴ you will learn that it is hard for government to buy expensive items because it must follow an elaborate policy of describing in great, mind-numbing detail what it wants to purchase (often without any helpful contact with the firms that sell these items), and then it must advertise for bids. Usually it has to accept the lowest price it is offered, even if this bid comes from a firm that is less competent than one that charges more.

For example, when a public agency wants to buy a computer, it cannot do so in the same way that business firms do. The latter will discuss computers with their manufacturers and get help in finding or designing the right machine. Often they will buy what they need from a firm they know and trust without ever describing the computer they want, sometimes paying a higher price in order to get the right combination of product and service. But this is not an option for government agencies; indeed, even attempting to do so would invite legal action. In such circumstances, as discussed later in the book, an agency’s purchasing agent may be accused of collusion, favoritism, unfairness, and sweetheart deals. Thus it has been next to impossible for the government purchaser to take into account its past experience with a supplier or that firm’s intangible qualities. Instead, the agency must produce an elaborate report describing what it wants and then accept sealed bids, with the sale going to the lowest bidder. It is hardly surprising that the government often winds up with second-rate computers sold by unhelpful firms.

Buying things is not a minor problem for Washington. Each year (as of the late 1990s) it spends about $200 billion purchasing items from business firms. That amount is almost one-seventh of the whole federal budget.

During the 1990s, efforts were made to improve government’s approach to purchasing. Steven Kelman, a scholar from whose research I took the computer story, was appointed by President Clinton to run the Office of Federal Procurement Policy (OFPP), part of the Office of Management and Budget. Kelman’s attempt to improve purchasing was supported by the NPR staff. He helped persuade Congress to pass two important bills, the Federal Acquisition Streamlining Act of 1994 and the Federal Acquisition Reform Act of 1995.⁵ Because of these laws and Kelman’s bureaucratic nudging, several changes ensued. The law now permitted agencies to buy off the shelf, without elaborate contracting, things that cost less than $100,000. Many agency purchasing officers began using credit cards to buy routine supplies from businesses, just as households do, thereby avoiding the need to fill out complicated purchase orders. And several agencies agreed to take into consideration the past performance of business firms in deciding from whom to buy materials. (Of course, neglected firms could still complain, but an agency was able to defend itself by making a written record of its reasons for having passed them over.)

At the same time, officials in the Defense Department, on their own initiative, began to cut back on the amount of detail required in requests for new military hardware—emphasizing, instead, performance goals for the hardware. The Pentagon probably did the most to change its procurement policies, because it stood to gain the most. Its budget was being sharply cut at the time; so, if it was going to pay for the new equipment it wanted, it would have to stretch its procurement dollars. Buying things more simply and cheaply meant that it could buy more of what it needed. (Inasmuch as Defense officials began to change how it dealt with contractors, what appears here in this book is now a bit out of date.)

But the problems of procurement reform remain, whatever the laws may say or agencies may require. Kelman found that the key problem was implementation—that is, persuading officials to buy things in a more streamlined way. Some agencies continued to believe that buying things the new, improved way would only get them into trouble—a not unreasonable fear given the fact that generations of civil servants had gotten into trouble for appearing to favor certain suppliers or failing to draw up lengthy purchasing specifications. And in any case, many of the limits on procurement were still the law of the land. Legislation continues to require that agencies favor buying things from American firms, help out small businesses, and give special status to firms headed by women, handicapped persons, disabled veterans, and racial minorities.

What is remarkable is the fact that, despite these fears and laws, many government officials are trying to be more efficient about how they buy things. The legal system is part of the constraints on government officials, but so is the culture within which laws work. Only time will tell whether this culture will change over the long haul.

Personnel

The NPR also wanted to reform federal personnel policies. In 1993, the head of the Office of Personnel Management (OPM) said he was for “radical change,”⁶ and the NPR has supported this view with ideas for letting agencies run their own personnel systems, making government employees accountable for achieving results rather than for following procedures, and encouraging entrepreneurial behavior. To signify the changes they wanted, President Clinton abolished the entire Federal Personnel Manual, the hopelessly long, intellectually arid legacy of all the rules that personnel specialists had accumulated over decades of effort.

But not much has happened in terms of how people are actually hired and managed. Indeed, little exists to take the place of the now-defunct personnel manual, leaving in place thousands of personnel experts who had grown up using the manual and its rules.⁷ OPM was supposed to abandon its traditional role as a centralized enforcer of civil service rules and embrace instead a support role in which it helped agency heads recruit their own workers. The NPR also urged that the fifteen pay bands that cover most federal employees be changed into a smaller number of pay levels, with agencies given more freedom in passing out wages based on agency needs and employee abilities.

But legislation that embodies these new approaches or alters the mind-set of some personnel officers has, as yet, been impossible to pass. Several bills—the Civil Service Simplification Act, the Federal Personnel Systems Reinvention Act, and the Omnibus Civil Service Reform Act—all failed to become law. If enacted, these laws would have decentralized personnel matters to each government agency, sharply reducing the old powers of OPM.

Not only did the bills fail to become law, but the civil service itself has been getting smaller. Federal employment shrank by about 12 percent in the Clinton years (with the biggest loser being the Defense Department). Some of these reductions were obtained through the buy-out of federal employees not yet ready for retirement. But the effect of this buy-out was to make the federal work force older, depriving it of many of the younger people in line to become top executives.

Nonetheless, some changes were made. The Federal Aviation Administration (FAA) was exempted from many old civil service rules so that it would be better able to recruit qualified air traffic controllers. The great public criticism of the Internal Revenue Service (IRS) led to some steps that may help its personnel policies. And in some agencies, improvements have been made in the management of personnel.

But it is still hard to get able people to work for the federal government, even though nearly 70 percent of all new workers were hired without having to go through the old system of centrally administered tests and hiring. Federal pay scales are now often lower than private ones, competition is fierce for highly skilled talent, and the government must worry about how many ethnic and racial minorities it hires. The personnel culture itself has not changed, however; as Patricia W. Ingraham reports, federal agencies had become accustomed to the rules, so they have gone on following them.⁸

The pilot projects for improving federal management described in this book here to here, especially the effort to improve hiring and paying workers at the China Lake Naval Air Station, still have not been as influential as some people hoped. As the book originally suggested, Congress and other federal agencies have not been very sympathetic to what was tried at China Lake and a few other locations.

President Clinton attempted a new approach by creating a number of “reinvention laboratories” in several agencies. These were rather informal efforts to use employee suggestions to find ways of abandoning silly rules. (In one case, a laboratory figured out how to allow employees to charge on an agency credit card small amounts of money—say, $25 or $30—without getting an elaborate written approval.) These laboratories may yet turn out to be a source of important change, but it is too early to tell.

The central reason why it is so difficult to change the ways in which people are hired and managed can be found in the contradiction built into our view of civil service. As Ingraham points out, these laws have always had a split personality: “[I]t is their responsibility to protect the permanent career work force from undue political influence, while still providing political leaders with a responsive set of career employees.”⁹ Balancing political independence with political accountability is very difficult, whatever the NPR may say about employee “empowerment.” The reality, as Paul Light puts it, is that rules will dominate: The government will “limit bureaucratic discretion through compliance with tightly drawn rules and regulations.”¹⁰

Making the Bureaucracy Perform Better

In 1993 Congress passed and the president signed the Government Performance and Results Act, a law designed to make it easier to link individual performance, agency accomplishments, and social goals. One of its mechanisms is evaluation of employee performance.

But how can this evaluation be achieved? The great majority of employees are now given a very high rating because there are few objective ways to link what a worker does and what an agency accomplishes.¹¹ In a business firm, the pay an employee gets is set by management, not by a government pay scale, and (when there is an incentive compensation plan) it is often based on his or her performance coupled, in many cases, with the performance of the firm as a whole. If the firm does well (in profits, market share, or stock price), employees are eligible for pay awards; the exact size of the award will, in theory, be based on each employee’s contribution to the firm’s success. In a government agency, by contrast, pay is largely set by a legislative scale and there are few, if any, measures of how well the agency as a whole has done. As a result, government performance evaluations tend to be very high because the managers who make these evaluations know that there are few, if any, objective grounds for making them lower. And employees would surely complain if, without such grounds, they were given a low rating; it would be taken as a sign of favoritism toward some friend of the boss who got a higher rating.

There will continue to be pushes for better performance, given what has happened with respect to procurement and personnel, this book tells a story of bureaucracy that need not be changed in any fundamental way from the story it told when it was first published.

The NPR wishes to create a government that costs less; but real cost savings—savings of the sort that can be seen by a taxpayer—can be achieved not by making marginal improvements in government management but, rather, by cutting the programs that government manages. Yet voters, although they want the government in general to spend less, also want more money spent on a wide range of programs from crime fighting and drug abuse to education, civil rights, and the environment. In the 1950s, none of these programs made up a significant part of the federal budget; what the government did then was limited to rather simple things—delivering mail, building highways, acquiring parks, conducting research, providing subsidies to various interests, and defending the nation. Today, it would be fairly easy to measure agency performance by how well the mail is delivered, how costly the highways and parks are, and how quickly the subsidies are delivered. But it is next to impossible to measure how well any agency reduces drug abuse, enhances civil rights, or educates children.

The NPR also would like to cut red tape; but red tape, as Herbert Kaufman has pointed out, is usually produced by organized political groups seeking to protect their interests by constraining how the government is managed.¹² To many groups, less red tape means fewer benefits. For example, lots of groups want better highways, but by better they mean not only well constructed and inexpensive but built in ways that serve other interests. Consider the following example: In 1956 the Federal Aid Highway Act created the interstate highway system with a law that was only twenty-eight pages long. Thirty-five years later, that program was reauthorized by the Intermodal Surface Transportation Efficiency Act, but the new law was 293 pages long. The difference in length can be explained by the multiplication of constraints: By 1991 we not only wanted to build more highways, we wanted to build them in ways that would aid mass transit, reduce air pollution, encourage the use of seat belts and motorcycle helmets, preserve historic sites, control erosion and outdoor advertising, use recycled rubber in making asphalt, buy iron and steel from U.S. manufacturers, define women as disadvantaged individuals, and protect Native American reservations—among other things. You can look it up. But I doubt the 1991 highway bill has empowered employees or made managing them any easier.

Clearly, public confidence in government is much lower now than it was when government was doing fewer things. No doubt there are many reasons for this—the rise of an adversarial media, the impact of Vietnam and Watergate, and an increase in certain social problems. But one additional reason may be that we have less confidence in government precisely because it is trying to do many things that cannot be done, by anyone, very well.

And even when it does things that can be done well, government often suffers by comparison with competitive business firms—not because the firms hire better people but because they are competing for customers instead of serving (captive) citizens. The NPR would like to make the government customer-friendly, and no doubt it may succeed in getting some agencies to distribute better information, provide clearer guidance to government procedures, and answer the telephone after only three rings. It would be nice to be treated as well by a government agency as one is by minimum-wage employees at McDonald’s or Burger King. But having visited both McDonald’s and my local post office in the last few days, I am not optimistic that the NPR—or any government effort—will make a big difference.

After reading this long recitation of the constraints under which government works, you make think that I am pessimistic about our political regime. Not at all. Every democratic government has these problems, and our government does better than most others in attending to these defects and trying to fix a few of the worst ones. In fact, by the standards of most democratic governments, many agencies in this country are friendlier and more cooperative than their counterparts abroad. Our constitutional system so fragments authority and encourages intervention that it produces two opposing bureaucratic effects: citizen-serving agencies that are friendlier and more responsive, and citizen-regulating agencies that are more rigid and adversarial.

The book you are about to read is an effort to set forth what is generally true about bureaucracy. Although it was published ten years ago, none of its central themes needs to be revised. Of course, the finer points of bureaucracy’s operations have changed, as evidenced by the ongoing struggles over procurement and personnel. And in time, the NPR may leave an important legacy of change. But for now we can say only that it is too early to determine the NPR’s actual effect. In short, even if some details are altered by this and other efforts, the broad picture remains the same.

J.Q.W.

PREFACE

THIS BOOK is an effort to explain why government agencies—bureaucracies—behave as they do. Though there may be lessons here for those interested in the behavior of government agencies abroad, the chief focus is on the United States. Toward the end of the book I offer a chapter’s worth of speculation as to differences between American and foreign bureaucracies.

It is a book filled with details about police departments, school systems, the CIA, the military, the State Department, regulatory commissions, the Postal Service, the Social Security Administration, the Army Corps of Engineers, the Forest Service, and many others. The details are not there simply to make the book long, but rather to persuade the reader that bureaucracy is not the simple, uniform phenomenon it is sometimes made out to be. Reality often does not conform to scholarly theories or popular prejudices.

Historians and sociologists interested in the evolution of the modern state have had to come to grips with the bureaucratization of society. Max Weber was the founder of this tradition and, though many of his specific claims have been disproved, he still repays close reading. But such a reading will convey a view of bureaucracy as a monolith—a distinctive form of social organization which exists to increase the predictability of government action by applying general rules to specific cases. Its members possess the authority of office, enjoy lifelong careers and high social esteem, and operate the levers of power in a way that makes the bureaucracy an overtowering force against which citizens and politicians often struggle in vain.

This view is partly correct, but accepting it without substantial modification will leave the reader ill-equipped to understand some important features of American bureaucracies (and perhaps bureaucracies generally): Not only do many agencies fail to apply general rules to specific cases, they positively resist any effort to set forth their policies in the form of clear and general rules. In increasing numbers of agencies, members whose authority rests on the offices they hold are being eclipsed in power and significance by members whose authority derives from the professional training they received. The supposedly overtowering power of bureaucracies is hard to reconcile with the ease with which legislatures in this country challenge, rebuke, and otherwise influence these agencies. Bureaucratic action is sometimes regular and predictable, but just as often it is irregular and unpredictable. Some bureaucracies resist change, but others seem always to be engaged in reorganizing their structures and modifying their doctrines. Max Weber’s insights are useful, but only up to a point. Beyond that point, distinctions are more important than generalizations.

Economists and political scientists have begun to apply to government agencies the same analytical methods that once were used to explain the behavior of business firms. Just as entrepreneurs are thought to be maximizing their “utility,” bureaucrats are now thought to be maximizing theirs. The utility of a business person is assumed to be profits; that of a bureaucrat is assumed to be something akin to profits: salary, rank, or power. Since both bureaucrats and business executives are people, it makes sense to assume that they prefer more of whatever they like to less. Some interesting theories of bureaucratic behavior have been produced by making this reasonable assumption, but the theories so far have not explained very much. I suspect one of the reasons is that bureaucrats have a variety of preferences; only part of their behavior can be explained by assuming they are struggling to get bigger salaries or fancier offices or larger budgets. Besides, bureaucrats don’t have as much control over their own salaries, offices, and budgets as do business people. For these reasons, or others yet to be discovered, it should not be surprising to find government agencies that actually refuse to take on new tasks or try to give up tasks they now perform.

Citizens and taxpayers have their own global view of bureaucracy. To them, bureaucrats are lethargic, incompetent hacks who spend their days spinning out reels of red tape and reams of paperwork, all the while going to great lengths to avoid doing the job they were hired to do. Their agencies chiefly produce waste, fraud, abuse, and mismanagement. That this view is an exaggeration is readily shown by public-opinion surveys in which people are asked about their personal experiences with government agencies. The great majority of the respondents say that these experiences were good, that the agency personnel were helpful, friendly, and competent. ¹ This can only mean that those lazy, incompetent bureaucrats must work for some other agency—the one the citizen never sees. But it is not quite clear what that other agency is. No one can seem to identify it, but everybody is sure it must exist. Maybe it is the one that buys $435 hammers. But on close inspection, even that turns out to be a bit of an exaggeration—there wasn’t a $435 hammer, though it is probably too late to convince anybody of that. (In chapter 17, I will try.)

Students of public administration have produced detailed studies of countless government agencies that, if one were to sit down and read them all, would put this into perspective. These books would show that some bureaucrats are hacks but most are not; that there is a lot of self-serving activity in agencies but also a lot that seems genuinely directed toward larger goals; and that some agencies work like Weberian bureaucracies, producing and applying rules, and some do not. But not many people are going to sit down and read them all. I have tried to. This book summarizes what I found.

It draws on over a quarter of a century of teaching and research on bureaucratic organizations, on the books and articles that others have written, and (especially) on the doctoral dissertations and seminar papers written by graduate students who have worked with me over the years. This academic knowledge has been supplemented by personal experience—as a naval officer for three years and as a part-time advisor to or member of various presidential commissions. Though I have learned a great deal from these encounters, nothing in this book consists of “inside dope” not already available in published sources. I dislike the kiss-and-tell memoirs that have recently become so popular—and so profitable. Many of them dishonor the public trust with which their authors were once endowed. From time to time, there will be unattributed statements about the workings of the FBI, the CIA, the State Department, and the Drug Enforcement Administration. These derive (unless otherwise noted) from my personal observations in an advisory capacity. Any scholar who studies these organizations will, I think, quickly learn the same thing without the benefit of privileged access.