The Rise of the Bureaucratic State
JAMES Q. WILSON
During its first 150 years, the American republic was not thought to have a "bureaucracy," and thus it would have been meaningless to refer to the problems" of a "bureaucratic state." There were, of course, appointed civilian officials. Though only about 3,000 at the end of the Federalist period, there were about 95,000 by the time Grover Cleveland assumed office in 1881, and nearly half a million by 1925. Some aspects of these numerous officials were regarded as problems—notably, the standards by which they were appointed and the political loyalties to which they were held—but these were thought to be matters of proper character and good management. The great political and constitutional struggles were not over the power of the administrative apparatus, but over the power of the President, of Congress, and of the states.
The Founding Fathers had little to say about the nature or function of the executive branch of the new government. The Constitution is virtually silent on the subject and the debates in the Constitutional Convention are almost
devoid of reference to an administrative apparatus. This reflected no lack of concern about the matter, however. Indeed, it was in part because of the Founders’ depressing experience with chaotic and inefficient management under the Continental Congress and the Articles of Confederation that they had assembled in Philadelphia. Management by committees composed of part-time amateurs had cost the colonies dearly in the War of Independence and few, if any, of the Founders wished to return to that system. The argument was only over how the heads of the necessary departments of government were to be selected, and whether these heads should be wholly subordinate to the President or whether instead they should form some sort of council that would advise the President and perhaps share in his authority. In the end, the Founders left it up to Congress to decide the matter.
There was no dispute in Congress that there should be executive departments, headed by single appointed officials, and, of course, the Constitution specified that these would be appointed by the President with the advice and consent of the Senate. The only issue was how such officials might be removed. After prolonged debate and by the narrowest of majorities, Congress agreed that the President should have the sole right of removal, thus confirming that the infant administrative system would be wholly subordinate—in law at least—to the President. Had not Vice President John Adams, presiding over a Senate equally divided on the issue, cast the deciding vote in favor of presidential removal, the administrative departments might conceivably have become legal dependencies of the legislature, with incalculable consequences for the development of the embryonic government.
THE "BUREAUCRACY PROBLEM"
The original departments were small and had limited duties. The State Department, the first to be created, had but nine employees in addition to the Secretary. The War Department did not reach 80 civilian employees until 1801; it commanded only a few thousand soldiers. Only the Treasury Department had substantial powers—it collected taxes, managed the public debt, ran the national bank, conducted land surveys, and purchased military supplies. Because of this, Congress gave the closest scrutiny to its structure and its activities.
The number of administrative agencies and employees grew slowly but steadily during the 19th and early 20th centuries and then increased explosively on the occasion of World War I, the Depression, and World War II. It is difficult to say at what point in this process the administrative system became a distinct locus of power or an independent source of political initiatives and problems. What is clear is that the emphasis on the sheer size of the administrative establishment—conventional in many treatments of the subject—is misleading.
The government can spend vast sums of money—wisely or unwisely— without creating that set of conditions we ordinarily associate with the bureaucratic state. For example, there could be massive transfer payments made under government auspices from person to person or from state to state, all managed by a comparatively small staff of officials and a few large computers. In 1971, the federal government paid out $54 billion under various social insurance programs, yet the Social Security Administration employs only 73,000 persons, many of whom perform purely routine tasks.
And though it may be harder to believe, the government could in principle employ an army of civilian personnel without giving rise to those organizational patterns that we call bureaucratic. Suppose, for instance, that we as a nation should decide to have in the public schools at least one teacher for every two students. This would require a vast increase in the number of teachers and school rooms, but almost all of the persons added would be performing more or less identical tasks, and they could be organized into very small units (e.g., neighborhood schools). Though there would be significant overhead costs, most citizens would not be aware of any increase in the "bureaucratic" aspects of education—indeed, owing to the much greater time each teacher would have to devote to each pupil and his or her parents, the citizenry might well conclude that there actually had been a substantial reduction in the amount of "bureaucracy."
To the reader predisposed to believe that we have a "bureaucracy problem," these hypothetical cases may seem farfetched. Max Weber, after all, warned us that in capitalist and socialist societies alike, bureaucracy was likely to acquire an "overpowering" power position. Conservatives have always feared bureaucracy, save perhaps the police. Humane socialists have frequently been embarrassed by their inability to reconcile a desire for public control of the economy with the suspicion that a public bureaucracy may be as immune to democratic control as a private one. Liberals have equivocated, either dismissing any concern for bureaucracy as reactionary quibbling about social progress, or embracing that concern when obviously nonreactionary persons (welfare recipients, for example) express a view toward the Department of Health, Education, and Welfare indistinguishable from the view businessmen take of the Internal Revenue Service.
There are at least three ways in which political power may be gathered undesirably into bureaucratic hands: by the growth of an administrative apparatus so large as to be immune from popular control, by placing power over a governmental bureaucracy of any size in private rather than public hands, or by investing discretionary authority in the hands of a public agency so that the exercise of that power is not responsive to the public good. These are not the only problems that arise because of bureaucratic organization. From the point of view of their members, bureaucracies are sometimes uncaring, ponderous, or unfair; from the point of view of their political superiors, they are some times unimaginative or inefficient; from the point of view of their clients, they are sometimes slow or unjust. No single account can possibly treat all that is problematic in bureaucracy; even the part I discuss here—the extent to which political authority has been transferred undesirably to an unaccountable administrative realm—is itself too large for a single essay. But it is, if not the most important problem, then surely the one that would most have troubled our Revolutionary leaders, especially those that went on to produce the Constitution. It was, after all, the question of power that chiefly concerned them, both in redefining our relationship with England and in finding a new basis for political authority in the Colonies.
To some, following in the tradition of Weber, bureaucracy is the inevitable consequence and perhaps necessary concomitant of modernity. A money economy, the division of labor, and the evolution of legal-rational norms to justify organizational authority require the efficient adaptation of means to ends and a high degree of predictability in the behavior of rulers. To this, Georg Simmel added the view that organizations tend to acquire the characteristics of those institutions, with which they are in conflict, so that as government becomes more bureaucratic, private organizations—political parties, trade unions, voluntary associations—will have an additional reason to become bureaucratic as well.
By viewing bureaucracy as an inevitable (or, as some would put it, "functional") aspect of society, we find ourselves attracted to theories that explain the growth of bureaucracy in terms of some inner dynamic to which all agencies respond and which makes all barely governable and scarcely tolerable. Bureaucracies grow, we are told, because of Parkinson's Law: Work and personnel expand to consume the available resources. Bureaucracies behave, we believe, in accord with various other maxims, such as the Peter Principle: In hierarchical organizations, personnel are promoted up to that point at which their incompetence becomes manifest—hence, all important positions are held by incompetents. More elegant, if not essentially different, theories have been propounded by scholars. The tendency of all bureaus to expand is explained by William A. Niskanen by the assumption, derived from the theory of the firm, that "bureaucrats maximize the total budget of their bureau during their tenure"—hence, "all bureaus are too large." What keeps them from being not merely too large but all-consuming is the fact that a bureau must deliver to some degree on its promised output, and if it consistently under delivers, its budget will be cut by unhappy legislators. But since measuring the output of a bureau is often difficult—indeed, even conceptualizing the output of the State Department is mind-boggling—the bureau has a great deal of freedom within which to seek the largest possible budget.
Such theories, both the popular and the scholarly, assign little importance to the nature of the tasks an agency performs, the constitutional framework in which it is embedded, or the preferences and attitudes of citizens and legislators. Our approach will be quite different: Different agencies will be examined in historical perspective to discover the kinds of problems, if any, to which their operation gave rise, and how those problems were affected— perhaps determined—by the tasks which they were assigned, the political system in which they operated, and the preferences they were required to consult. What follows will be far from a systematic treatment of such matters, and even farther from a rigorous testing of any theory of bureaucratization: Our knowledge of agency history and behavior is too sketchy to permit that.
BUREAUCRACY AND SIZE
During the first half of the 19th century, the growth in the size of the federal bureaucracy can be explained, not by the assumption of new tasks by the government or by the imperialistic designs of the managers of existing tasks, but by the addition to existing bureaus of personnel performing essentially routine, repetitive tasks for which the public demand was great and unavoidable. The principal problem facing a bureaucracy thus enlarged was how best to coordinate its activities toward given and noncontroversial ends.
The increase in the size of the executive branch of the federal government at this time was almost entirely the result of the increase in the size of the Post Office. From 1816 to 1861, federal civilian employment in the executive branch increased nearly eight fold (from 4,837 to 36,672), but 86 percent of this growth was the result of additions to the postal service. The Post Office Department was expanding as population and commerce expanded. By 1869 there were 27,000 post offices scattered around the nation; by 1901, nearly 77,000. In New York alone, by 1894 there were nearly 3,000 postal employees, the same number required to run the entire federal government at the beginning of that century.
The organizational shape of the Post Office was more or less fixed in the administration of Andrew Jackson. The Postmaster General, almost always appointed because of his partisan position, was aided by three (later four) assistant postmaster generals dealing with appointments, mail-carrying contracts, operations, and finance. There is no reason in theory why such an organization could not deliver the mails efficiently and honestly: The task is routine, its performance is measurable, and its value is monitored by millions of customers. Yet the Post Office, from the earliest years of the 19th century, was an organization marred by inefficiency and corruption. The reason is often thought to be found in the making of political appointments to the Post Office. "Political hacks," so the theory goes, would inevitably combine dishonesty and incompetence to the disservice of the nation; thus, by cleansing the department of such persons these difficulties could be avoided. Indeed, some have argued that it was the advent of the "spoils system" under Jackson that contributed to the later inefficiencies of the public bureaucracy.
The opposite is more nearly the case. The Jacksonians did not seek to make the administrative apparatus a mere tool of the Democratic party advantage, but to purify that apparatus not only of what they took to be Federalist subversion but also of personal decadence. The government was becoming not just large, but lax. Integrity and diligence were absent, not merely from government, but from social institutions generally. The Jacksonians were in many cases concerned about the decline in what the Founders had called "republican virtue," but what their successors were more likely to call simplicity and decency. As Matthew Crenson has recently observed in his book The Federal Machine, Jacksonian administrators wanted to "guarantee the good behavior of civil servants" as well as to cope with bigness, and to do this they sought both to place their own followers in office and—what is more important—to create a system of depersonalized, specialized bureaucratic rule. Far from being the enemies of bureaucracy, the Jacksonians were among its principal architects.
Impersonal administrative systems, like the spoils system, were "devices for strengthening the government's authority over its own civil servants"; these bureaucratic methods were, in turn, intended to "compensate for a decline in the disciplinary power of social institutions" such as the community, the professions, and business. If public servants, like men generally in a rapidly growing and diversifying society, could no longer be relied upon "to have a delicate regard for their reputations," accurate bookkeeping, close inspections, and regularized procedures would accomplish what character could not.
Amos Kendall, Postmaster General under President Jackson, set about to achieve this goal with a remarkable series of administrative innovations. To prevent corruption, Kendall embarked on two contradictory courses of action:
He sought to bring every detail of the department's affairs under his personal scrutiny and he began to reduce and divide the authority on which that scrutiny depended. Virtually every important document and many unimportant ones had to be signed by Kendall himself. At the same time, he gave to the Treasury Department the power to audit his accounts and obtained from Congress a law requiring that the revenues of the department be paid into the Treasury rather than retained by the Post Office. The duties of his subordinates were carefully defined and arranged so that the authority of one assistant would tend to check that of another. What was installed was not simply a
specialized management system, but a concept of the administrative separation of powers.
Few subsequent postmasters were of Kendall's ability. The result was predictable. Endless details flowed to Washington for decision but no one in Washington other than the Postmaster General had the authority to decide. Meanwhile, the size of the postal establishment grew by leaps and bounds. Quickly the department began to operate on the basis of habit and local custom: Since everybody reported to Washington, in effect no one did. As Leonard D. White was later to remark, "the system could work only because it was a vast, repetitive, fixed, and generally routine operation." John Wanamaker, an able businessman who became Postmaster General under President Cleveland, proposed decentralizing the department under 26 regional supervisors. But Wanamaker's own assistants in Washington were unenthusiastic about such a diminution in their authority and, in any event, Congress steadfastly refused to endorse decentralization.
Civil service reform was not strongly resisted in the Post Office; from 1883 on, the number of its employees covered by the merit system expanded. Big city postmasters were often delighted to be relieved of the burden of dealing with hundreds of place-seekers. Employees welcomed the job protection that civil service provided. In time, the merit system came to govern Post Office personnel almost completely, yet the problems of the department became, if anything, worse. By the mid-20th century, slow and inadequate service, an inability technologically to cope with the mounting flood of mail, and the inequities of its pricing system became all too evident. The problem with the Post Office, however, was not omnipotence but impotence. It was a government monopoly. Being a monopoly, it had little incentive to find the most efficient means to manage its services; being a government monopoly, it was not free to adopt such means even when found—communities, Congressmen, and special-interest groups saw to that.
THE MILITARY ESTABLISHMENT
Not all large bureaucracies grow in response to demands for service. The Department of Defense, since 1941 the largest employer of federal civilian officials, has become, as the governmental keystone of the "military-industrial complex," the very archetype of an administrative entity that is thought to be so vast and so well-entrenched that it can virtually ignore the political branches of government, growing and even acting on the basis of its own inner imperatives. In fact, until recently the military services were a major economic and political force only during wartime. In the late 18th and early 19th centuries, America was a neutral nation with only a tiny standing army during the Civil War, over two million men served on the Union side alone and the War Department expanded enormously, but demobilization after the war was virtually complete, except for a small Indian-fighting force. Its peacetime authorized strength was only 25,000 enlisted men and 2,161 officers, and its actual strength for the rest of the century was often less. Congress authorized the purchase and installation of over 2,000 coastal defense guns, but barely six percent of these were put in place.
When war with Spain broke out, the army was almost totally unprepared. Over 300,000 men eventually served in that brief conflict, and though almost all were again demobilized, the War Department under Elihu Root was reorganized and put on a more professionalized basis with a greater capacity for unified central control. Since the United States had become an imperial power with important possessions in the Caribbean and the Far East, the need for a larger military establishment was clear; even so, the average size of the army until World War I was only about 250,000.
The First World War again witnessed a vast mobilization—nearly five million men in all—and again an almost complete demobilization after the war The Second World War involved over 16 million military personnel. The demobilization that followed was less complete than after previous engagements, owing to the development of the Cold War, but it was substantial nonetheless—the Army fell in size from over eight million men to only half a million. Military spending declined from $91 billion in the first quarter of 1945 to only slightly more than $10 billion in the second quarter of 1947. For the next three years it remained relatively flat. It began to rise rapidly in 1950, partly to finance our involvement in the Korean conflict and partly to begin the construction of a military force that could counterbalance the Soviet Union, especially in Europe.
In sum, from the Revolutionary War to 1950, a period of over 170 years, the size and deployment of the military establishment in this country was governed entirely by decisions made by political leaders on political grounds. The military did not expand autonomously, a large standing army did not find wars to fight, and its officers did not play a significant potential role except in wartime and occasionally as presidential candidates. No bureaucracy proved easier to control, at least insofar as its size and purposes were concerned.
A "MILITARY-INDUSTRIAL COMPLEX"?
The argument for the existence of an autonomous, bureaucratically led military industrial complex is supported primarily by events since 1950. Not only has the United States assumed during this period worldwide commitments that necessitate a larger military establishment, but the advent of new, high-technology weapons has created a vast industrial machine with an interest in sustaining a high level of military expenditures, especially on weapons research, development, and acquisition. This machine, so the argument goes, is allied with the Pentagon in ways that dominate the political officials nominally in charge of the armed forces. There is some truth in all this. We have become a world military force, though that decision was made by elected officials in 1949-1950 and not dictated by a (then nonexistent) military-industrial complex. High cost, high technology weapons have become important and a number of industrial concerns will prosper or perish depending on how contracts for those weapons are let. The development and purchase of weapons is sometimes made in a wasteful, even irrational, manner. And the allocation of funds among the several armed services is often dictated as much by interservice rivalry as by strategic or political decisions.
But despite all this, the military has not been able to sustain itself at its preferred size, to keep its strength constant or growing, or to retain for its use a fixed or growing portion of the Gross National Product. Even during the last two decades, the period of greatest military prominence, the size of the Army has varied enormously—from over 200 maneuver battalions in 1955, to 174 in 1965, rising to 217 at the peak of the Vietnam action in 1969, and then declining rapidly to 138 in 1972. Even military hardware, presumably of greater interest to the industrial side of the military-industrial complex, has often declined in quantity, even though per unit price has risen. The Navy had over 1,000 ships in 1955; it has only 700 today. The Air Force had nearly 24,000 aircraft in 1955; it has fewer than 14,000 today. This is not to say the combat strength of the military is substantially less than it once was, and there is greater firepower now at the disposal of each military unit, and there are various missile systems now in place, for which no earlier counterparts existed. But the total budget, and thus the total force level, of the military has been decided primarily by the President and not in any serious sense forced upon him by subordinates. (For example, President Truman decided to allocate one third of the federal budget to defense, President Eisenhower chose to spend no more than 10 percent of the Gross National Product on it, and President Kennedy strongly supported Robert McNamara's radical and controversial budget revisions.) Even a matter of as great significance as the size of the total military budget for research and development has proved remarkably resistant to inflationary trends: In constant dollars, since 1964 that appropriation has been relatively steady (in 1972 dollars, about $30 billion a year).
The principal source of growth in the military budget in recent years has arisen from congressionally determined pay provisions. The legislature has voted for more or less automatic pay increases for military personnel with the result that the military budget has gone up even when the number of personnel in the military establishment has gone down.
The bureaucratic problems associated with the military establishment arise mostly from its internal management and are functions of its complexity, the uncertainty surrounding its future deployment, conflicts among its constituent services over mission and role, and the need to purchase expensive equipment without the benefit of a market economy that can control costs. Complexity, uncertainty, rivalry, and monopsony are inherent (and frustrating) aspects of the military as a bureaucracy, but they are very different problems from those typically associated with the phrase "the military-industrial complex." The size and budget of the military are matters wholly within the power of civilian authorities to decide—indeed, the military budget contains the largest discretionary items in the entire federal budget.
If the Founding Fathers were to return to review their handiwork, they would no doubt be staggered by the size of both the Post Office and the Defense Department, and in the case of the latter, be worried about the implications of our commitments to various foreign powers. They surely would be amazed at the technological accomplishments but depressed by the cost and inefficiency of both departments; but they would not, I suspect, think that our Constitutional arrangements for managing these enterprises have proved defective or that there had occurred, as a result of the creation of these vast bureaus, an important shift in the locus of political authority.
They would observe that there have continued to operate strong localistic pressures in both systems—offices are operated, often uneconomically, in some small communities because small communities have influential congressmen; military bases are maintained in many states because states have powerful senators. But a national government with localistic biases is precisely the system they believed they had designed in 1787, and though they surely could not have then imagined the costs of it, they just as surely would have said (Hamilton possibly excepted) that these costs were the defects of the system's virtues.
BUREAUCRACY AND CLIENTELISM
After 1861, the growth in the federal administrative system could no longer be explained primarily by an expansion of the postal service and other traditional bureaus. Though these continued to expand, new departments were added that reflected a new (or at least greater) emphasis on the enlargement of the scope of government. Between 1861 and 1901, over 200,000 civilian employees were added to the federal service, only 52 percent of whom were postal workers. Some of these, of course, staffed a larger military and naval establishment stimulated by the Civil War and the Spanish-American War. By 1901 there were over 44,000 civilian defense employees, mostly workers in government-owned arsenals and shipyards. But even these could account for less than one fourth of the increase in employment during the preceding 40 years.
What was striking about the period after 1861 was that the government began to give formal, bureaucratic recognition to the emergence of distinctive interests in a diversifying economy. As Richard L. Schott has written, "whereas earlier federal departments had been formed around specialized governmental functions (foreign affairs, war, finance, and the like), the new departments of this period—Agriculture, Labor, and Commerce—were devoted to the interests and aspirations of particular economic groups."
The original purpose behind these clientele-oriented departments was neither to subsidize nor to regulate, but to promote, chiefly by gathering and publishing statistics and (especially in the case of agriculture) by research. The formation of the Department of Agriculture in 1862 was to become a model, for better or worse, for later political campaigns for government recognition. A private association representing an interest—in this case the United States Agricultural Society—was formed. It made every President from Fillmore to Lincoln an honorary member, it enrolled key congressmen, and it began to lobby for a new department. The precedent was followed by labor groups, especially the Knights of Labor, to secure creation in 1888 of a Department of Labor. It was broadened in 1903 to be a Department of Commerce and Labor, but 10 years later, at the insistence of the American Federation of Labor, the parts were separated and the two departments we now know were formed.
There was an early l9th-century precedent for the creation of these client serving departments: the Pension Office, then in the Department of the Interior. Begun in 1833 and regularized in 1849, the Office became one of the largest bureaus of the government in the aftermath of the Civil War, as hundreds of thousands of Union Army veterans were made eligible for pensions if they had incurred a permanent disability or injury while on military duty; dependent widows were also eligible if their husbands had died in service or of service-connected injuries. The Grand Army of the Republic (GAR), the leading veterans' organization, was quick to exert pressure for more generous pension laws and for more liberal administration of such laws as already existed. In 1879 congressmen, noting the number of ex-servicemen living (and voting) in their states, made veterans eligible for pensions retroactively to the date of their discharge from the service, thus enabling thousands who had been late in filing applications to be rewarded for their dilatoriness. In 1890 the law was changed again to make it unnecessary to have been injured in the service—all that was necessary was to have served and then to have acquired a permanent disability by any means other than through "their own vicious habits." And whenever cases not qualifying under existing law came to the attention of Congress, it promptly passed a special act making those persons eligible by name.
So far as is known, the Pension Office was remarkably free of corruption in the administration of this windfall—and why not, since anything an administrator might deny, a legislator was only too pleased to grant. By 1891 the Commissioner of Pensions observed that his was "the largest executive bureau in the world." There were over 6,000 officials supplemented by thousands of local physicians paid on a fee basis. In 1900 alone, the Office had to process 477,000 cases. Fraud was rampant as thousands of persons brought false or exaggerated claims; as Leonard D. White was later to write, "pensioners and their attorneys seemed to have been engaged in a gigantic conspiracy to defraud their own government." Though the Office struggled to be honest, Congress was indifferent—or more accurately, complaisant: The GAR was a powerful electoral force and it was ably and lucratively assisted by thousands of private pension attorneys. The pattern of bureaucratic clientelism was set in a way later to become a familiar feature of the governmental landscape—a subsidy was initially provided, because it was either popular or unnoticed, to a group that was powerfully benefited and had few or disorganized opponents; the beneficiaries were organized to supervise the administration and ensure the funding of the program; the law authorizing the program, first passed because it seemed the right thing to do, was left intact or even expanded because politically it became the only thing to do. A benefit once bestowed cannot easily be withdrawn.
PUBLIC POWER AND PRIVATE INTERESTS
It was at the state level, however, that client-oriented bureaucracies proliferated in the l9th century. Chief among these were the occupational licensing agencies. At the time of Independence, professions and occupations either could be freely entered (in which case the consumer had to judge the quality of service for himself) or entry was informally controlled by the existing members of the profession or occupation by personal tutelage and the management of reputations. The latter part of the l9th century, however, witnessed the increased use of law and bureaucracy to control entry into a line of work. The state courts generally allowed this on the grounds that it was a proper exercise of the "police power" of the state, but as Morton Keller has observed, "when state courts approved the licensing of barbers and blacksmiths, but not of horse shoers, it was evident that the principles governing certification were—to put it charitably—elusive ones." By 1952, there were more than 75 different occupations in the United States for which one needed a license to practice, and the awarding of these licenses was typically in the hands of persons already in the occupation, who could act under color of law. These licensing boards—for plumbers, dry cleaners, beauticians, attorneys, undertakers, and the like—frequently have been criticized as particularly flagrant examples of the excesses of a bureaucratic state. But the problems they create—of restricted entry, higher prices, and lengthy and complex initiation procedures—are not primarily the result of some bureaucratic pathology but of the possession of public power by persons who use it for private purposes. Or more accurately, they are the result of using public power in ways that benefited those in the profession in the sincere but unsubstantiated conviction that doing so would benefit the public generally.
The New Deal was perhaps the high water mark of at least the theory of bureaucratic clientelism. Not only did various sectors of society, notably agriculture, begin receiving massive subsidies, but the government proposed, through the National Industrial Recovery Act (NRA), to cloak with public power a vast number of industrial groupings and trade associations so that they might control production and prices in ways that would end the depression. The NRA's Blue Eagle fell before the Supreme Court—the wholesale delegation of public power to private interests was declared unconstitutional. But the piecemeal delegation was not, as the continued growth of specialized promotional agencies attests. The Civil Aeronautics Board, for example, erroneously thought to be exclusively a regulatory agency, was formed in 1938 "to promote" as well as to regulate civil aviation and it has done so by restricting entry and maintaining above-market-rate fares.
Agriculture, of course, provides the leading case of clientelism. Theodore J. Lowi finds "at least 10 separate, autonomous, local self-governing systems" located in or closely associated with the Department of Agriculture that control to some significant degree the flow of billions of dollars in expenditures and loans. Local committees of farmers, private farm organizations, agency heads, and committee chairmen in Congress dominate policy-making in this area—not, perhaps, to the exclusion of the concerns of other publics, but certainly in ways not powerfully constrained by them.
The growing edge of client-oriented bureaucracy can be found, however, not in governmental relations with private groups, but in the relations among governmental units. In dollar volume, the chief clients of federal domestic expenditures are state and local government agencies. To some degree, federal involvement in local affairs by the cooperative funding or management of local enterprises has always existed. The Northwest Ordinance of 1784 made public land available to finance local schools and the Morrill Act of 1862 gave land to support state colleges, but what Morton Grodzins and Daniel Elazar have called "cooperative federalism," though it always existed, did not begin in earnest until the passage in 1913 of the 16th Amendment to the Constitution allowed the federal government to levy an income tax on citizens and thereby to acquire access to vast sources of revenue. Between 1914 and 1917, federal aid to states and localities increased a thousand fold. By 1948 it amounted to over one tenth of all state and local spending; by 1970, to over one sixth.
The degree to which such grants, and the federal agencies that administer them, constrain or even direct state and local bureaucracies is a matter of dispute. No general answer can be given—federal support of welfare programs has left considerable discretion in the hands of the states over the size of benefits and some discretion over eligibility rules, whereas federal support of highway construction carries with it specific requirements as to design, safety, and (since 1968) environmental and social impact.
A few generalizations are possible, however. The first is that the states and not the cities have been from the first, and remain today, the principal client group for grants-in-aid. It was not until the Housing Act of 1937 that money was given in any substantial amount directly to local governments, and though many additional programs of this kind were later added, as late as 1970 less than 12 percent of all federal aid went directly to cities and towns. The second general observation is that the 1960s mark a major watershed in the way in which the purposes of federal aid are determined. Before that time, most grants were for purposes initially defined by the states—to build highways and airports, to fund unemployment insurance programs, and the like. Beginning in the 1960s, the federal government, at the initiative of the President and his advisors, increasingly came to define the purposes of these grants— not necessarily over the objection of the states, but often without any initiative from them. Federal money was to be spent on poverty, ecology, planning, and other "national" goals for which, until the laws were passed, there were few, if any, well-organized and influential constituencies. Whereas federal money was once spent in response to the claims of distinct and organized clients, public or private, in the contemporary period federal money has increasingly been spent in ways that have created such clients.
And once rewarded or created, they are rarely penalized or abolished. What David Stockman has called the "social pork barrel" grows more or less steadily. Between 1950 and 1970, the number of farms declined from about 5.6 million to fewer than three million, but government payments to farmers rose from $283 million to $3.2 billion. In the public sector, even controversial programs have grown. Urban renewal programs have been sharply criticized,
but federal support for the program rose from $281 million in 1965 to about $1 billion in 1972. Public housing has been enmeshed in controversy, but federal support for it rose from $206 million in 1965 to $845 million in 1972.
Federal financial support for local poverty programs under the Office of Economic Opportunity has actually declined in recent years, but this cut is almost unique and it required the steadfast and deliberate attention of a determined President who was bitterly assailed both in the Congress and in the courts.
If the Founding Fathers were to return to examine bureaucratic clientelism, they would, I suspect, be deeply discouraged. James Madison clearly foresaw that American society would be "broken into many parts, interests and classes of citizens" and that this "multiplicity of interests" would help ensure against "the tyranny of the majority," especially in a federal regime with separate branches of government. Positive action would require a "coalition of a major ity"; in the process of forming this coalition, the rights of all would be protected, not merely by self-interested bargains, but because in a free society such a coalition "could seldom take place on any other principles than those of justice and the general good." To those who wrongly believed that Madison thought of men as acting only out of base motives, the phrase is instructive: Persuading men who disagree to compromise their differences can rarely be achieved solely by the parceling out of relative advantage; the belief is also required that what is being agreed to is right, proper, and defensible before public opinion.
Most of the major new social programs of the United States, whether for the good of the few or the many, were initially adopted by broad coalitions appealing to general standards of justice or to conceptions of the public weal. This is certainly the case with most of the New Deal legislation—notably such programs as Social Security—and with most Great Society legislation—notably Medicare and aid to education; it was also conspicuously the case with respect to post-Great Society legislation pertaining to consumer and environmental concerns. State occupational licensing laws were supported by majorities interested in, among other things, the contribution of these statutes to public safety and health.
But when a program supplies particular benefits to an existing or newly created interest, public or private, it creates a set of political relationships that make exceptionally difficult further alteration of that program by coalitions of the majority. What was created in the name of the common good is sustained in the name of the particular interest. Bureaucratic clientelism becomes self perpetuating, in the absence of some crisis or scandal, because a single interest group to which the program matters greatly is highly motivated and well situated to ward off the criticisms of other groups that have a broad but weak interest in the policy.
In short, a regime of separated powers makes it difficult to overcome objections and contrary interests sufficiently to permit the enactment of a new program or the creation of a new agency. Unless the legislation can be made to pass either with little notice or at a time of crisis or extraordinary majorities and sometimes even then—the initiation of new programs requires public interest arguments. But the same regime works to protect agencies, once created, from unwelcome change because a major change is, in effect, new
legislation that must overcome the same hurdles as the original law, but this time with one of the hurdles—the wishes of the agency and its client—raised much higher. As a result, the Madisonian system makes it relatively easy for the delegation of public power to private groups to go unchallenged and, therefore, for factional interests that have acquired a supportive public bureaucracy to rule without submitting their interests to the effective scrutiny
and modification of other interests.
BUREAUCRACY AND DISCRETION
For many decades, the Supreme Court denied to the federal government any general "police power" over occupations and businesses, and thus most such regulation occurred at the state level and even there under the constraint that it must not violate the notion of "substantive due process"—that is, the view that there were sharp limits to the power of any government to take (and therefore to regulate) property. What clearly was within the regulatory province of the federal government was interstate commerce, and thus it is not surprising that the first major federal regulatory body should be the Interstate Commerce Commission (ICC), created in 1887.
What does cause, if not surprise, then at least dispute, is the view that the Commerce Act actually was intended to regulate railroads in the public interest. It has become fashionable of late to see this law as a device sought by the railroads to protect themselves from competition. The argument has been given its best-known formulation by Gabriel Kolko. Long-haul railroads, facing ruinous price wars and powerless to resist the demands of big shippers
for rebates, tried to create voluntary cartels or "pools" that would keep rates high. These pools always collapsed, however, when one railroad or another would cut rates in order to get more business. To prevent this, the railroads
turned to the federal government seeking a law to compel what persuasion could not induce. But the genesis of the act was in fact more complex: Shippers wanted protection from high prices charged by railroads that operated
monopolistic services in certain communities; many other shippers served by competing lines wanted no legal barriers to prevent competition from driving prices down as far as possible; some railroads wanted regulation to ease competition, while others feared regulation. And the law as finally passed in fact made "pooling'' (or cartels to keep prices up) illegal.
The true significance of the Commerce Act is not that it allowed public power to be used to make secure private wealth but that it created a federal commission with broadly delegated powers that would have to reconcile con-
flicting goals (the desire for higher or lower prices) in a political environment characterized by a struggle among organized interests and rapidly changing technology. In short, the Commerce Act brought forth a new dimension to
the problem of bureaucracy: not those problems, as with the Post Office, that resulted from size and political constraints, but those that were caused by the need to make binding choices without any clear standards for choice.
The ICC was not, of course, the first federal agency with substantial discretionary powers over important matters. The Office of Indian Affairs, for a while in the War Department but after 1849 in the Interior Department, coped for the better part of a century with the Indian problem equipped with no clear policy, beset on all sides by passionate and opposing arguments, and infected with a level of fraud and corruption that seemed impossible to eliminate. There were many causes of the problem, but at root was the fact that the government
was determined to control the Indians but could not decide toward what end that control should be exercised (extermination, relocation, and assimilation all had their advocates) and, to the extent the goal was assimilation, could find no method by which to achieve it. By the end of the century, a policy of relocation had been adopted de facto and the worst abuses of the Indian service had been eliminated—if not by administrative skill, then by the exhaustion of things in Indian possession worth stealing. By the turn of the century, the management
of the Indian question had become the more or less routine administration of Indian schools and the allocation of reservation land among Indian claimants.
REGULATION VERSUS PROMOTION
It was the ICC and agencies and commissions for which it was the precedent that became the principal example of federal discretionary authority. It is important, however, to be clear about just what this precedent was. Not everything we now call a regulatory agency was in fact intended to be one. The ICC, the Antitrust Division of the Justice Department, the Federal Trade Commission (FTC), the Food and Drug Administration (FDA), the National Labor Relations Board (NLRB)—all these were intended to be genuinely regulatory bodies created to handle under public auspices matters once left to private arrangements. The techniques they were to employ varied: approving rates (ICC), issuing cease-and-desist orders (FTC), bringing civil or criminal actions in the courts (the Antitrust Division), defining after a hearing an appropriate standard of conduct (NLRB), or testing a product for safety (FDA). In each case, however, Congress clearly intended that the agency either define its own standards (a safe drug, a conspiracy in restraint of trade, a fair labor practice) or choose among competing claims (a higher or lower rate for shipping grain).
Other agencies often grouped with these regulatory bodies—the Civil Aeronautics Board, the Federal Communications Commission, the Maritime Commission—were designed, however, not primarily to regulate, but to promote the development of various infant or threatened industries. However, unlike fostering agriculture or commerce, fostering civil aviation or radio broadcasting was thought to require limiting entry (to prevent "unsafe" aviation or broadcast interference); but at the time these laws were passed few believed that the restrictions on entry would be many, or that the choices would be made on any but technical or otherwise noncontroversial criteria. We smile now at their naivete, but we continue to share it—today we sometimes suppose that choosing an approved exhaust emission control system or a water pollution control system can be done on the basis of technical criteria and without affecting production and employment.
The creation of regulatory bureaucracies has occurred, as is often remarked, in waves. The first was the period between 1887 and 1890 (the Commerce Act and the Antitrust Act), the second between 1906 and 1915 (the Pure Food and Drug Act, the Meat Inspection Act, the Federal Trade Commission Act, the Clayton Act), the third during the 1930s (the Food, Drug, and Cosmetic Act, the Public Utility Holding Company Act, the Securities Exchange Act, the Natural Gas Act, the National Labor Relations Act), and the fourth during the latter part of the 1960s (the Water Quality Act, the Truth in Lending Act, the National Traffic and Motor Vehicle Safety Act, various amendments to the drug laws, the Motor Vehicle Pollution Control Act, and many others).
Each of these periods was characterized by progressive or liberal Presidents in office (Cleveland, T. R. Roosevelt, Wilson, F. D. Roosevelt, Johnson); one was a period of national crisis (the 1930s); three were periods when the President enjoyed extraordinary majorities of his own party in both houses of Congress (1914-1916, 1932-1940, and 1964-1968); and only the first period preceded the emergence of the national mass media of communication. These facts are important because of the special difficulty of passing any genuinely regulatory legislation: A single interest, the regulated party, sees itself seriously threatened by a law proposed by a policy entrepreneur who must appeal to an unorganized majority, the members of which may not expect to be substantially or directly benefited by the law. Without special political circumstances—a crisis, a scandal, extraordinary majorities, an especially vigorous President, the support of media—the normal barriers to legislative innovation (i.e., to the formation of a "coalition of the majority") may prove insuperable.
Stated another way, the initiation of regulatory programs tends to take the form of majoritarian rather than coalitional politics. The Madisonian system is placed in temporary suspense: Exceptional majorities propelled by a public mood and led by a skillful policy entrepreneur take action that might not be possible under ordinary circumstances (closely divided parties, legislative executive checks and balances, popular indifference). The consequence of majoritarian politics for the administration of regulatory bureaucracies is great. To initiate and sustain the necessary legislative mood, strong, moralistic, and sometimes ideological appeals are necessary—leading, in turn, to the granting of broad mandates of power to the new agency (a modest delegation of authority would obviously be inadequate if the problem to be resolved is of crisis proportions), or to the specifying of exacting standards to be enforced (e.g., no carcinogenic products may be sold, 95 percent of the pollutants must be eliminated), or to both.
Either in applying a vague but broad rule ("the public interest, convenience, and necessity") or in enforcing a clear and strict standard, the regulatory agency will tend to broaden the range and domain of its authority, to lag behind technological and economic change, to resist deregulation, to stimulate corruption, and to contribute to the bureaucratization of private institutions.
It will broaden its regulatory reach out of a variety of motives: to satisfy the demand of the regulated enterprise that it be protected from competition, to make effective the initial regulatory action by attending to the unanticipated side effects of that action, to discover or stretch the meaning of vague statutory language, or to respond to new constituencies induced by the existence of the agency to convert what were once private demands into public pressures. For example, the Civil Aeronautics Board, out of a desire both to promote aviation and to protect the regulated price structure of the industry, will resist the entry into the industry of new carriers. If a Public Utilities Commission sets rates too low for a certain class of customers, the utility will allow service to those customers to decline in quality, leading in turn to a demand that the Commission also regulate the quality of service. If the Federal Communications Commission cannot decide who should receive a broadcast license by applying the "public interest" standard, it will be powerfully tempted to invest that phrase with whatever preferences the majority of the Commission then entertains, leading in turn to the exercise of control over many more aspects of broadcasting than merely signal interference—all in the name of deciding what the standard for entry shall be. If the Antitrust Division can prosecute conspiracies in restraint of trade, it will attract to itself the complaints of various firms about business practices that are neither conspiratorial nor restraining but merely competitive, and a "vigorous" antitrust lawyer may conclude that these practices warrant prosecution.
Regulatory agencies are slow to respond to change for the same reason all organizations with an assured existence are slow: There is no incentive to respond. Furthermore, the requirements of due process and of political conciliation will make any response time-consuming For example, owing to the complexity of the matter and the money at stake, any comprehensive review of the long-distance rates of the telephone company will take years, and possibly, decades.
Deregulation, when warranted by changed economic circumstances or undesired regulatory results, will be resisted. Any organization, and a fortiori and public organization, develops a genuine belief in the rightness of its mission that is expressed as a commitment to regulation as a process. This happened to the ICC in the early decades of this century as it steadily sought both enlarged powers (setting minimum as well as maximum rates) and a broader jurisdiction (over trucks, barges, and pipelines as well as railroads). It even urged incorporation into the Transportation Act of 1920 language directing it to prepare a comprehensive transportation plan for the nation. Furthermore, any regulatory agency will confer benefits on some group or interest, whether intended or not; those beneficiaries will stoutly resist deregulation. (But in happy proof of the fact that there are no iron laws, even about bureaucracies, we note the recent proposals emanating from the Federal Power Commission that the price of natural gas be substantially deregulated.)
The operation of regulatory bureaus may tend to bureaucratize the private sector. The costs of conforming to many regulations can be met most easily—often, only—by large firms and institutions with specialized bureaucracies of their own. Smaller firms and groups often must choose between unacceptably high overhead costs, violating the law, or going out of business. A small bakery producing limited runs of a high-quality product literally may not be able to meet the safety and health standards for equipment, or to keep track of and administer fairly its obligations to its two employees; but unless the bakery is willing to break the law, it must sell out to a big bakery that can afford to do these things, but may not be inclined to make and sell good bread.
I am not aware of any data that measure private bureaucratization or industrial concentration as a function of the economies of scale produced by the need to cope with the regulatory environment, but I see no reason why
such data could not be found.
Finally, regulatory agencies that control entry, fix prices, or substantially affect the profitability of an industry create a powerful stimulus for direct or indirect forms of corruption. The revelations about campaign finance in the 1972 presidential election show dramatically that there will be a response to that stimulus. Many corporations, disproportionately those in regulated industries (airlines, milk producers, oil companies), made illegal or hard-to-justify campaign contributions involving very large sums.
THE ERA OF CONTRACT
It is far from clear what the Founding Fathers would have thought of all this. They were not doctrinaire exponents of laissez-faire, nor were 18th-century governments timid about asserting their powers over the economy. Every
imaginable device of fiscal policy was employed by the states after the Revolutionary War. Mother England had, during the mercantilist era, fixed prices and wages, licensed merchants, and granted monopolies and subsidies.
(What were the royal grants of American land to immigrant settlers but the greatest of subsidies, sometimes—as in Pennsylvania—almost monopolistically given?) European nations regularly operated state enterprises, controlled
trade, and protected industry. But as William D. Gramp.p has noted, at the Constitutional Convention the Founders considered authorizing only four kinds of economic controls, and they rejected two of them. They agreed to
allow the Congress to regulate international and interstate commerce and to give monopoly protection in the form of copyrights and patents. Even Madison's proposal to allow the federal government to charter corporations was
rejected. Not one of the 85 Federalist papers dealt with economic regulation; indeed, the only reference to commerce was the value to it of a unified nation and a strong navy.
G. Warren Nutter has speculated as to why our Founders were so restrained in equipping the new government with explicit regulatory powers. One reason may have been the impact of Adam Smith's Wealth of Nations, published the same year as the Declaration of Independence, and certainly soon familiar to many rebel leaders, notably Hamilton. Smith himself sought to explain the American prosperity before the Revolution by the fact that Britain, through "salutary neglect," had not imposed mercantilist rules on the colonial economy. "Plenty of good land, and liberty to manage their own affairs in their own way" were the "two great causes" of colonial prosperity. As Nutter observes, there was a spirit of individualistic venture among the colonies that found economic expression in the belief that voluntary contracts were the
proper organization principle of enterprise.
One consequence of this view was that the courts in many states were heavily burdened with cases testing the provisions of contracts and settling debts under them. In one rural county in Massachusetts the judges heard over
800 civil cases during 1785. As James Willard Hurst has written, the years before 1875 were "above all else, the years of contract in our law."
The era of contract came to an end with the rise of economic organizations so large or with consequences so great that contracts were no longer adequate, in the public's view, to adjust corporate behavior to the legitimate expecta-
tions of other parties. The courts were slower to accede to this change than were many legislatures, but in time they acceded completely, and the era of administrative regulation was upon us. The Founders, were they to return, would understand the change in the scale and social significance of enterprise, would approve of many of the purposes of regulation, perhaps would approve of the behavior of some of the regulatory bureaus seeking to realize those pur-
poses, but surely would be dismayed at the political cost resulting from having vested vast discretionary authority in the hands of officials whose very existence—to say nothing of whose function—was not anticipated by the Con-
stitutional Convention, and whose effective control is beyond the capacity of the governing institutions which that Convention had designed.
THE BUREAUCRATIC STATE AND THE REVOLUTION
The American Revolution was not only a struggle for independence but a fundamental rethinking of the nature of political authority. Indeed, until that reformulation was completed the Revolution was not finished. What made political authority problematic for the colonists was the extent to which they believed Mother England had subverted their liberties despite the protection of the British constitution, until then widely regarded in America as the most perfect set of governing arrangements yet devised. The evidence of usurpation is now familiar: unjust taxation, the weakening of the independence of the judiciary, the stationing of standing armies, and the extensive use of royal patronage to reward office-seekers at colonial expense. Except for the issue of taxation, which raised for the colonists major questions of representation, almost all of their complaints involved the abuse of administrative powers.
The first solution proposed by Americans to remedy this abuse was the vesting of most (or, in the case of Pennsylvania and a few other states, virtually all) powers in the legislature. But the events after 1776 in many colonies, notably Pennsylvania, convinced the most thoughtful citizens that legislative abuses were as likely as administrative ones: In the extreme case, citizens would suffer from the "tyranny of the majority." Their solution to this problem was, of course, the theory of the separation of powers by which, as brilliantly argued in The Federalist papers, each branch of government would check the likely usurpations of the other.
This formulation went essentially unchallenged in theory and unmodified by practice for over a century. Though a sizeable administrative apparatus had come into being by the end of the 19th century, it constituted no serious
threat to the existing distribution of political power because it either performed routine tasks (the Post Office) or dealt with temporary crises (the military). Some agencies wielding discretionary authority existed, but they either dealt with groups whose liberties were not of much concern (the Indian Office) or their exercise of discretion was minutely scrutinized by Congress (the Land Office, the Pension Office, the Customs Office). The major discretionary agencies of the l9th century flourished at the very period of greatest congressional domination of the political process—the decades after the Civil War—and thus, though their supervision was typically inefficient and sometimes corrupt, these agencies were for most practical purposes direct dependencies of Congress. In short, their existence did not call into question the theory of the separation of powers.
But with the growth of client-serving and regulatory agencies, grave questions began to be raised—usually implicitly—about that theory. A client serving bureau, because of its relations with some source of private power, could become partially independent of both the executive and legislative branches—or in the case of the latter, dependent upon certain committees and independent of others and of the views of the Congress as a whole. A regulatory agency (that is to say, a truly regulatory one and not a clientelist or promotional agency hiding behind a regulatory fig leaf) was, in the typical case, placed formally outside the existing branches of government. Indeed, they were called "independent" or "quasi-judicial" agencies (they might as well have been called "quasi-executive" or "quasi-legislative") and thus the special status that clientelist bureaus achieved de facto, the regulatory ones achieved de jure.
It is, of course, inadequate and misleading to criticize these agencies, as has often been done, merely because they raise questions about the problem of sovereignty. The crucial test of their value is their behavior, and that can be judged only by applying economic and welfare criteria to the policies they produce. But if such judgments should prove damning, as increasingly has been the case, then the problem of finding the authority with which to alter or abolish such organizations becomes acute. In this regard the theory of the separation of powers has proved unhelpful.
The separation of powers makes difficult, in ordinary times, the extension of public power over private conduct—as a nation, we came more slowly to the welfare state than almost any European nation, and we still engage in less central planning and operate fewer nationalized industries than other democratic regimes. But we have extended the regulatory sway of our national government as far or farther than that of most other liberal regimes (our environmental and safety codes are now models for much of Europe), and the bureaus wielding these discretionary powers are, once created, harder to change or redirect than would be the case if authority were more centralized.
The shift of power toward the bureaucracy was not inevitable. It did not result simply from increased specialization, the growth of industry, or the imperialistic designs of the bureaus themselves. Before the second decade of this century, there was no federal bureaucracy wielding substantial discretionary powers. That we have one now is the result of political decisions made by elected representatives. Fifty years ago, the people often wanted more of government than it was willing to provide—it was, in that sense, a republican government in which representatives moderated popular demands. Today, not only does political action follow quickly upon the stimulus of public interest, but government itself creates that stimulus and sometimes acts in advance of it.
All democratic regimes tend to shift resources from the private to the public sector and to enlarge the size of the administrative component of government. The particularistic and localistic nature of American democracy has created a particularistic and client-serving administration. If our bureaucracy often serves special interests and is subject to no central direction, it is because our legislature often serves special interests and is subject to no central leadership. For Congress to complain of what it has created and it maintains is, to be charitable, misleading. Congress could change what it has devised, but there is little reason to suppose it will.