The Online Library of Liberty

A project of Liberty Fund, Inc.

• Preface to the Second Edition
• One: the Doctrine of the Separation of Powers and Institutional Theory
• Two: the Foundation of the Doctrine
• Three: the Theory of the Balanced Constitution
• Four: Montesquieu
• Five: the Matchless Constitution and Its Enemies
• Six: the Doctrine In America
• Seven: the Doctrine In France
• Eight: the Rise and Fall of Parliamentary Government
• Nine: From the Third Republic to the Fifth
• Ten: Progressivism and Political Science In America
• Eleven: Political Theory, Constitutionalism, and the Behavioural Approach
• Twelve: a Model of a Theory of Constitutionalism
• Thirteen: Epilogue: the Separation of Powers and the Administrative State
• Bibliography

THIRTEEN

Epilogue: The Separation of Powers and the Administrative State

The Separation of Powers and Political Theory

When the first edition of this book was published in 1967 it was extremely unfashionable. The history and analysis of an institutional theory concerned with the limitation of the power of government clashed with the dominant intellectual trends of the time. There were then three powerful streams of thought, all of which, from very different standpoints, were hostile to the idea that there could be a coherent tradition of political thought about the institutional structure of government that had something useful to say about the way in which government impacted upon the liberty of the citizen.

The first, Marxism, saw political institutions as the instruments of the domination of the proletariat by the bourgeoisie, a mere reflection of the economic relations of production, the differing details of the structure of which in different capitalist countries could have no more than a trivial effect upon political outcomes. Thirty years on, this line of thought has little to tell us about the study of politics. The utopian assumptions of Marxist thought about human nature, so far from the realism of Montesquieu or Madison, can provide no viable basis for political analysis.

There were, however, two other intellectual tendencies, one predominantly American, the other predominantly British, in origin, each equally dismissive of constitutional or institutional theory. Behaviourism was then at its high watermark and is dealt with at some length in Chapter 11. Essentially behaviourism propounded a demand for empirical verification in terms which were impossible to realize. But the problems which had for centuries been the concern of theorists dealing with the separation of powers and other institutional safeguards did not go away and became in some respects more acute. The inability of behaviourists to address these problems effectively has now become fully apparent.

The other attack came from the work of those whom, for want of a better term, I shall describe as the sceptics. The sceptics found it impossible to see how theories could be framed in a manner which could generate valid propositions across space and time. On the one hand, Peter Winch attacked the concept of social science as a comparative study.1 On the other hand, Quentin Skinner denied the very possibility of writing a meaningful history of a concept such as the separation of powers.2 Skinner went so far as to say that “it must be a mistake even to try either to write intellectual biographies concentrating on the works of a given writer, or to write histories of ideas tracing the morphology of a given concept over time. Both these types of study . . . are necessarily misconceived.”3 He argued that “the classic texts are concerned with their own quite alien problems. . . .”4 Because, according to this view, it was impossible for us to understand properly the meaning of texts from earlier periods of history, he concluded that “any statement . . . is inescapably the embodiment of a particular intention, on a particular occasion, addressed to the solution of a particular problem, and thus specific to its situation in a way that it can only be naive to try to transcend.”5

Ten years later Skinner had abandoned this untenable line of argument. In a book on the history of political thought he wrote, “I have thus tried to write a history centred less on the classic texts and more on the history of ideologies, my aim being to construct a general framework within which the writings of the more prominent theorists can be situated.”6 The importance of the mistake made by Skinner in his earlier work is that he misunderstood the essential continuity of human thought, the extent to which one writer builds upon the work of another, even if only by reacting against it. It is impossible to draw hard and fast divisions between periods of thought and to put them into watertight compartments. Earlier thinkers were not a series of different species, they were human, as we are. The contextual details were different, to be sure, a fact we must always be aware of; but the problems, the concerns, and the dilemmas were essentially the same as those we face today.

These attacks upon a tradition of thought stretching back to Aristotle were perhaps doomed to fail; but the reason that the theory of the separation of powers remains significant, however, is because the problem it addresses is as salient now as it was in the seventeenth or eighteenth century. Far from its central issue having changed fundamentally, the nature of the political problem that concerned earlier writers on the separation of powers has remained exactly the same: how to control the power of government. The rise of the administrative state, the weakening of the effective power of legislatures, and the problem of democratic control—these are the new concerns, but they are variations on an old theme. The theory of the separation of powers is an empirical theory. It embodies values, but in a hypothetical sense. “If you wish to safeguard liberty, then. . . .” The object of this Epilogue must therefore be to ask where political thought stands today in relation to the theory of the separation of powers. We shall survey the main areas of interest that have evolved over the past thirty years in Britain and the United States and then attempt to draw together the threads of the argument into a credible theory of political institutions consonant with the model developed in Chapter 12, above.

The Problem of Government: The United States

In both the United States and Great Britain the concern with the practical problems of the articulation of government and its impact upon the rights and liberties of the individual have been prominent themes in public life and academic literature alike over the past thirty years. Since the 1970’s the problem of the separation of powers has exercised the American courts more than at any time in their history. This concern was evidenced in a wide variety of fields, and the issues involved were not trivial; they went to the heart of the problems of modern government. The output of books and articles on the separation of powers during this period bears witness to the importance attached to the issues that are at stake.7

The court order requiring President Nixon to produce the Watergate tapes in evidence raised in a stark form the question of the extent to which the Chief Executive was entitled to the privilege of confidentiality for his communications with his advisers. In United States v. Nixon the Supreme Court accepted, some would say invented,8 the right of the President to withhold information from Congress, holding this privilege to be “fundamental to the operation of Government and inextricably rooted in the separation of powers under the Constitution.”9 Nevertheless, the claim of confidentiality could not “prevail over the fundamental demands of due process of law in the fair administration of criminal justice.”10 The Court ordered the tapes to be produced, and the resignation of the President became inevitable.

Another decision of the Court, less dramatic perhaps but potentially more wide-ranging, invalidated the use of the “legislative veto,” a device which had been developed to enable Congress to control the way in which the Administration carried out the laws which the Congress had enacted. By 1983 Congress had inserted nearly two hundred legislative vetoes into statutes, making it possible to strike down specific actions of the executive branch. When legislation was passed delegating power to the executive, provisions were included in the legislation which allowed vetoes by one House of Congress, by both Houses, by a congressional committee, or even by a committee chair of decisions made by executive departments or independent agencies.11 In this way Congress could intervene in the administrative process to reverse decisions of which Congress, or in reality some congressmen or senators, disapproved. In the Chadha Case, in 1983,12 the Court considered the constitutionality of a legislative veto exercised by the House of Representatives reversing a decision of the Attorney General, which would have allowed Chadha, a student whose visa had expired, to remain in the United States. The Chief Justice, delivering the opinion of the Court declaring the legislative veto unconstitutional, argued that the Constitution had erected checks to the exercise of power by each branch of government, and that “to preserve those checks, and maintain the separation of powers, the carefully defined limits on the power of each Branch must not be eroded.”13 Although this case did not prevent Congress from making further use of the legislative veto,14 it did raise vitally important issues to which we will return.

In 1985 Congress passed the Gramm-Rudman-Hollings Act, the Balanced Budget and Emergency Control Act. The intention of the Act was to reduce the federal budget deficit to zero in the fiscal year 1991. Part of the process of achieving this aim was to give to the Comptroller General, an official appointed by the President with the approval of the Senate but removable only by a joint resolution of Congress or impeachment, the duty of reporting to the President his conclusions on the measures necessary to give effect to the legislation. The Supreme Court, in considering the validity of this legislation, based its decision on the doctrine of the separation of powers. “That this system of division and separation of powers produces conflicts, confusion, and discordance at times is inherent, but it was deliberately so structured to assure full, vigorous, and open debate on the great issues affecting the people and to provide avenues for the operation of checks on the exercise of governmental power.”15 The Court found that the Act charged the Comptroller General, an officer subject to removal by Congress, with making “executive” decisions, and that therefore the Congress “in effect has retained control over the execution of the Act and has intruded into the executive function. The Constitution does not permit such intrusion.”16

As a reaction to Richard Nixon’s dismissal of Watergate Special Prosecutor Archibald Cox, Congress passed the Ethics in Government Act of 1978, which was intended to limit the President’s control over such investigations. The Act provided for the judicial appointment of Independent Counsel to investigate alleged misdeeds of senior members of the executive branch.17 It had been argued before the Court that the Act infringed upon the separation of powers, because the function of prosecution, being essentially executive in nature, should not be entrusted to an officer with a degree of independence of the President. The Court rejected this view and upheld the constitutionality of the Act. The dissent of Justice Scalia in this case, however, and in Mistretta v. U.S. demands further examination.18

The question of the circumstances in which individuals can sue public bodies for redress, and therefore involve the courts in the detail of the regulatory process, provided another occasion for the Supreme Court to appeal to the doctrine of the separation of powers. In 1984 the Court refused standing to sue to a group of parents of black children who complained that tax exemptions granted by the IRS to racially discriminatory private schools impaired their ability to have public schools in their area desegregated.19 The Court argued that

Professor Sunstein has remarked that the Court’s attitude in this case reflects “a form of judicial skepticism about both government regulation and court entanglement in executive functions.”21 This scepticism lies “in the belief that administrative regulation, grounded as it is in technocratic expertise and political accountability, is incompatible with judicial oversight.”22

In a number of other cases the federal courts have used the separation of powers to resolve issues before them: the constitutionality of the Sentencing Reform Act of 1984;23 the use by the President of advisory committees,24 including his wife’s membership of one of them;25 the jurisdiction of the courts in international law26 and extradition cases;27 and in 1995 the Supreme Court’s invalidation of a section of a federal statute which required the courts to reopen certain cases on which final judgements had already been made.28

Clearly the Supreme Court has pursued its own agenda in trying to reconcile the problems of the modern State with its view of the intentions of the Founding Fathers. But whether one agrees with the Court’s decisions in individual cases, the over-all problem remains: the balance between the control that each branch of government exercises over the others and the degree of co-ordination among them necessary to the effective operation of government.29 The almost feverish activity of the American judiciary in recent years in attempting to draw this line reflects, in part at least, the political reality of “gridlock,” the fact that in only six of the thirty years from 1969 to 1998 will the presidency and a majority of both Houses of Congress have been in the hands of the same political party. Although the significance of party allegiance in the American system of government is not as great as in European states, this record of split control of the executive and legislative branches, culminating in the capture of both Houses by the Republicans in 1994 and the continuation of this situation in 1996, does reflect an important consequence of the American version of the separation of powers.30 The precise reasons why divided government became so significant after 1968 are not clear, and as the constitutional separation of powers existed both before and after this watershed it cannot be blamed, or credited, with the whole responsibility for this situation. However, as the mechanism which facilitates this American form of “coalition government” it is clearly a very significant factor.

For some this disjunction is simply the desirable realization of the Founders’ prescription of “weak government,”31 for others it is the cause of “the structural inability of our government to propose, legislate and administer a balanced program for governing.”32 There is little point in retracing in detail the outworn controversy over the proposals that have been made for the United States to adopt parliamentary government, but we shall return later to the implications for the United States of divided government.33 Although the British model of parliamentary government is by no means the sole version on offer,34 it is the one which has been promoted since Woodrow Wilson advocated it in 1879. The attractions of the “Westminster model,” however, have to be judged in terms of its performance in the last decade of the twentieth century, and it is to this issue we now turn.

Government by Party in Britain

In Chapter 12, above, the argument was developed that, although Britain did not have the kind of separation of powers that the Founding Fathers adopted in the United States, nevertheless the branches of government were not “fused” and there were still important boundaries between them. Today much of that argument still stands, but a number of factors have further eroded the values which lie behind the doctrine of the separation of powers and which have therefore led to the emasculation of the checks and balances that still, even after the Second World War, had some significance. In the mid-nineteenth century it was possible to see a balance between Parliament and Government such that neither dominated the other. Parliamentary government meant that the legislature could exercise control over the government by the use of the vote of confidence, or the vote of censure, and the government could restrain the excesses of the House of Commons by the threat of dissolution. In the period from 1832 to 1868, eight governments were defeated in the House of Commons and resigned immediately or obtained a dissolution and then were either beaten in the ensuing election or failed to win the confidence of the House after the election. Even when a government had a nominal majority in the House, the bonds of party were too weak to enable it to discipline its “supporters” if they disliked the party’s policies. This was a balanced system of government, operating within the framework of a very limited franchise and reflecting the views of a middle-class electorate still dominated by a landowning aristocracy. It was the model of parliamentary government which became so admired in other countries, but the “balance” of legislative and executive powers it represented rested upon a very special set of social and political conditions which were soon to change.

The extensions of the franchise in 1867 and 1884 ushered in a period of cabinet government in which the executive, because of the strengthening of party ties, began to dominate the legislature, and the likelihood of a government defeat in the House of Commons (except when a minority government was in office) grew smaller and smaller.35 The rise of the Labour Party, with its strong ideological basis, accelerated these tendencies, and with the Labour victory of 1945, Parliament became the forum for set battles between Government and Opposition, in which, because of the tight discipline exercised by the party leaders, there was no real possibility that the legislature, as such, would have a role to play. Whilst a government with even a slender majority in the House of Commons was in office decision-making was not carried on in the legislature. Policy decisions were made in government departments, in cabinet committees, in the offices of cabinet ministers, and above all in the office of the Prime Minister. The legislature retained its importance, because its procedures imposed upon the government the necessity of presenting and defending its measures in public. During the legislative process, however, no more than marginal changes in government proposals were made, and then usually because the government had become aware of faults in its own drafting or changed its mind on some point of detail on which its own party members felt particularly strongly. The justification of the system depended on the concept of the united party, sustained by an ideology and a sense of loyalty to its principles. If a united party had a majority in the Commons, the fact that the Opposition could make little or no difference to policy outcomes did not of itself destroy the concept of rule by the majority of the legislature. The major parties connived, and still do connive, in the maintenance of this fiction, for it was assumed that each would get its turn, and none wanted to spoil its future exercise of power. In reality there have always been internal party divisions on policy, which are resolved by discussion or by the threat of the exercise of party discipline, or both. In fact it is within the ruling party that decisions are taken, by the party, and not by the legislature. This then is the system of “government by party,” which although subject to certain restraints, in particular the concern with the outcome of the next election, can no longer be described, by any stretch of the imagination, as “parliamentary government.”36

Under the prime ministership of John Major government by party took a new twist which effectively made the workings of the system quite apparent. Under previous regimes, both Conservative and Labour, the constitutional myths of the system, ministerial responsibility and the collective responsibility of government to Parliament, which were the basis of the “accountability” of government, had been sedulously maintained. However, the long period of office of the Conservative Party, since 1979, and the looser style of leadership of John Major made it very apparent that the locus of decision-taking was well outside Parliament. Mr. Major’s perceived weakness compared with his predecessor, Margaret Thatcher, well-publicised differences within his cabinet, and the intensification of divisions within the Conservative Party over European integration led to the emergence of a large group of dissidents in the Conservative parliamentary party who were prepared to challenge Major’s leadership to the point of voting against him in the House of Commons, provided that they did not actually bring down the Government.

Faced with defeat on a number of issues the Government made it clear that it did not consider resignation necessary if defeated in the Commons, even on an important matter of policy. The Government would resign only if defeated on a formal vote of confidence.37 Thus party dissidents felt relatively free to harass their leaders in the Commons, openly fighting battles over policy, voting against government measures, going to the brink with threats to bring down the Government, but drawing back at the last moment.

In 1995 Britain was subjected to the ultimate expression of the operation of government by party. The Prime Minister, beset by Europhobe rebels, both in the cabinet and more generally in the parliamentary party, continually threatening to destroy him if he did not accede to their demands, resigned as leader of his party and stood for re-election as leader, as a means of reasserting his authority. There followed the amazing spectacle of a contest between two candidates for the leadership of the Conservative Party, one of whom happened to be the Prime Minister of the United Kingdom. That one political party, by its internal discussions and political manoeuvering, should determine government policy on vital national issues and decide who should hold the highest office in the land, while the rest of the country stood by powerless to intervene, was surely a travesty of “parliamentary government.” John Major won the contest, a personal triumph for him perhaps, but a death-knell for any concept of parliamentary government in Britain, for, if he had lost, the country would have had a new Prime Minister and a new set of policies, without a general election, as a result of a decision of the majority of the majority party, and without Parliament, as such, having had any say in the outcome.38

It has now become clear that the myths which sustain the constitutional theory of governmental accountability to the electorate are in fact the greatest obstacle to genuine accountability. The formal theory of ministerial responsibility, the requirement of the accountability of the executive to the legislature, in practice makes real accountability impossible to achieve. The reason is that criticism of the process of administration amounts to criticism of the government, of the ministers, and of their ability to govern. Therefore ministers resist any kind of investigation of alleged maladministration, they maintain a pall of secrecy over what goes on in the administrative machine, and because they are technically responsible for all that goes on in their departments, they defend the indefensible. The situation is made worse by the need of the majority party in the House to support the government, so that informed criticism is further suppressed. The Select Committee on the Public Service recognized this in 1996. “A Minister’s survival in his job depends primarily on the satisfaction of his ministerial colleagues—particularly the Prime Minister—and of his fellow Members of Parliament . . . so long as his ministerial and party colleagues are prepared to defend him, the chances of obtaining his removal are minimal.”39

To some extent this situation has changed in recent years. The government has in part privatised the administrative machine, and as a result refuses to accept responsibility for errors, as in the case of the Home Secretary and the Prison Service Agency. Without adequate controls over the privatised functions, however, and without clear boundaries between the responsibilities of ministers and agency heads, the minister is still seen as responsible, and indeed it remains open for ministers to intervene when it suits them to do so. The secrecy and impenetrability of the British system of government can therefore be clearly attributed to the limited extent of the application of the separation of powers in Britain. In the United States the absence of formal “responsible government,” the separation of powers, in fact, has made investigation and criticism of the administration much easier and more acceptable, making it possible to pass the Freedom of Information Act and to have regulatory authorities with teeth.

The complete failure of parliamentary control over the administrative machine in Britain is witnessed by the amazing proliferation of regulators and “ombudsmen” that has characterized the past thirty years. The privatization of public utilities has resulted in the creation of a series of pale imitations of the American regulatory commissions, in an attempt to give the impression that the interests of consumers were being protected. These bodies include the Director General of Telecommunications, the Director General of Gas Supply, the Director General of Water Supply, the Director General of Electricity Supply, the Director General of Fair Trading, and the Ombudsman for Legal Services. In the public sector there is the Parliamentary Commissioner for Administration, the Health Service Commissioner, the Northern Ireland Parliamentary Commissioner for Administration and Commissioner for Complaints, the Police Complaints Authority, the Independent Commission for Police Complaints in Northern Ireland, the Broadcast Complaints Commission, the Data Protection Registrar, the Council on Tribunals, the Local Government Ombudsman, the Prisons Ombudsman, among others. Nothing could be a clearer acknowledgment of the inability of Parliament to control the administrative machine than this plethora of independent regulatory officials, but the government’s acceptance of this fact for the purposes of public relations did not lead them to set up an effective control over administrative agencies. The powers of the ombudsmen, at any rate those who deal with the public sector, are inadequate. They have little power to enforce their decisions, and their work is effectively advisory rather than regulatory. Another indicator of the failure of the traditional mechanisms of control has been the appointment of extra-parliamentary committees, chaired by senior judges, to look into standards of public life (the Nolan Committee), and arms sales to Iraq (the Scott Committee).

Furthermore in Britain we have recently seen a most extraordinary misuse of the executive power over the legislature. In the Deregulation and Contracting Out Act of 1994, in the words of one commentator, “Part I . . . conferred upon Ministers a power to suspend any provision of an Act of Parliament if they were of the opinion that the effect of the provision in question was to impose a burden affecting any person in the carrying on of any trade. Part II of the Act provided for the transfer of statutory functions from the Ministers to which they were entrusted to private sector contractors at the discretion of the Minister in question.”40 This was not a delegation of legislative power, it was an abdication of the power of the legislature. This domination of government by party, however, is becoming increasingly out of step with trends in public opinion which reflect a loosening of old party allegiances and a greater independence of electoral behaviour.

The Administrative State

The modern world is characterized by the development of what has been described as the “administrative state.” Bureaucracy has been with us, of course, for more than two hundred years, but the modern administrative state exhibits such complexity of structures and such a proliferation of rules that the earlier conception of an “executive” consisting of a body of civil servants putting into effect, under the direction of ministers, the commands of the legislature is no longer tenable. The distinction between political leaders and bureaucrats has simultaneously become sharper and more confused. Sharper because the administrative state has taken on an autonomy of its own—it is only marginally under the control of its political masters at any point in time. A large proportion of a country’s budget is committed by existing legislation so that even a government determined on change can achieve very little in the short run, and the complexity of the rules governing economic and social life is such that even small changes can produce quite unexpected results. Thus the so-called executive, the political leaders nominally responsible to the legislature for the conduct of government, may in practice have little real control over the government machine, a fact which is only too obvious if we examine the attempts which have been made to develop new procedures for the control of the administration. The confusion arises because the political leaders spend a great deal of their time attempting to manage the administrative machinery for which they are ostensibly responsible but which they are invariably unable to control.

The reality of the working of government provides a difficulty for the traditional theory of the separation of powers, which divides the powers and functions of government into three, for, in effect, there are now four major sections of the political institutions of the democratic state: the legislature, the government, the administrative machine, and the judiciary. The term “executive” long ago lost its original connotation, because the political leaders at the head of the state machinery, whether the presidency or a cabinet in a parliamentary system, became deeply involved in the formulation and initiation of policy and legislation as well as their implementation. Now the label “executive” is even less appropriate, because it is the administrative machine, influenced but not controlled by the political leadership, which carries the laws into effect. It would be less confusing if we were able to drop the term “executive” altogether, a solution relatively easy in Britain, where the term “government” is normally used, but a very difficult solution in the United States, where the “Executive Power” is embedded in the Constitution. For analytical purposes, however, it would be more satisfactory to avoid using the term “executive,” and we will attempt to do so. Equally the term “government” is ambiguous, and we will therefore designate the political leadership as the “policy branch.”

The logic of the present situation, therefore, is to accept that there are now four branches of government, each with its own structures and processes, and to provide the control mechanisms necessary to prevent the abuse of power by any of them. Thus the attempt by the Supreme Court of the United States to define only three branches of government and to distribute functions among them is an impossible task, particularly, as we will argue, when their functional analysis is inadequate. This leaves open the question of the appropriate methods by which to control each branch, but does not, of course, imply that any one branch of government can be exclusively entrusted with only one function.

It is increasingly clear that the problem of administering the modern State is made more and more difficult by the assumption that the administrative machine should be under the direct control of the “government,” the policy branch. Politicians are in general incompetent to control that machinery, and indeed, by their meddling in the day-to-day working of the administration they create enormous problems, reacting to the political pressures of the moment. In Britain the uncertainty and confusion caused in this way, in such fields as education, transport, and health, are only too obvious. Almost weekly changes in policy imposed by weak and ineffectual ministers leave the administrative machine in chaos. Only too often in past years universities, schools, or hospitals have not been told the size of their annual budget until the financial year is already well advanced, because the policy branch could not resolve its internal conflicts.

In this respect the earlier advocates of administrative autonomy were correct.41 Unless the administrative state is abolished altogether—an unlikely eventuality—in some sense politics will have to be taken out of administration. This does not mean, however, that the administrative machine should be left to get on with the job uncontrolled, as some of those writers would suggest. It means that effective methods of control must be established to safeguard the rights of the individual to prevent the abuse of power by administrators, as much as by the legislature or the policy branch.

Judicial Review and the Administrative State

The failure of legislatures to exercise adequate controls over the administration has led, both in the United States and in Britain, to attempts by the judiciaries to fill this gap. The rise of the administrative state and the “Death of the Separation of Powers,” among other things, lead Gary Lawson to argue that the processes of judicial review since 1789 have created a situation in which “one cannot have allegiance both to the administrative state and to the Constitution.” If one then chooses the administrative state over the Constitution “all constitutional discourse is rendered problematic. . . . What is left of the Constitution after excision of its structural provisions, however interesting it may be as a matter of normative political theory, simply is not the Constitution.”42 It is not exactly clear what the practical results would be of accepting this “originalist” view, which presumably would return the American system of government to its position in 1789, but it does lead one to consider the way in which the Supreme Court has attempted to deal with the problem of the separation of powers, particularly over the past thirty years.

‡Professor William Gwyn has set out very clearly the major ideological rift between the Justices in the cases that came before the Supreme Court in the Reagan era.43 On the one hand, the “formalists” would wish to limit each branch of government to the exercise of a “power” given to it by the Constitution. On the other hand, the “functionalists” take a more flexible view, examining each situation where one branch is accused of having intruded on the “primary function” of another branch. One problem of this classification is that the so-called formalists, in order to maintain their position, have to attempt to define the nature of the “legislative power,” the “executive power,” and the “judicial power.” The ambiguity of the term “power” leads them in fact into a dependence on establishing the essential function of a particular branch of government. Thus Justice Scalia, the arch-formalist, declares that “governmental investigation and prosecution of crimes is a quintessentially executive function,”44 and that “all purely executive power must be under the control of the President.”45 It is not revealed how we are to determine the purity of the power in question. In fact problems met by the Supreme Court in attempting to define the “powers” or functions of government reflect very closely the chaotic discussion of this subject by the Committee on Ministers’ Powers in Britain in the 1930’s. It was clear many years ago that attempts to allocate particular functions precisely to particular branches of government must fail. It is possible to define four abstract functions—rule-making, a discretionary function, rule-application, and rule-adjudication46 —but quite impossible to allocate them exclusively to different branches of government, because all human behaviour involves all four functions to some degree. A partial resolution of this dilemma comes from Jesse Choper, who argues that the federal judiciary ought not to “decide constitutional questions concerning the respective powers of Congress and the President vis-à-vis one another,” but that the determination of these issues should be left to “the interplay of the national political process.”47 While this view would make many of the post-war issues that have come before the Supreme Court in the area of the separation of powers non-justiciable, it would still leave in place the problem of the role of the judiciary in regulating the administrative state, which is concerned more with the impact of administrative action upon individual rights and liberties than with the grand issues of “legislative-executive relations.” Certainly those who advocate judicial restraint in disputes between government and legislature have logic on their side, if for no other reason than that no sound functional basis can be found for making decisions in this area. In the absence of any other effective control over the administrative machine, however, the role of the courts in defending individual rights must remain. It is therefore very important to distinguish between actions of “the policy branch” and actions of the “administration,” if this control is to be exercised effectively.

In Britain the role of the judiciary has also been changing. The first edition of this work included a rather harsh judgement on the British judiciary and its “failure of nerve.” Traditionally the British courts have been extremely tolerant of the power of the “executive,” reflecting the dominant mythology of a government’s executing the will of a sovereign Parliament and aware that a government with a secure parliamentary majority could quickly reverse any decisions of the courts which were distasteful to it. In recent years, however, there has been a change in the attitude of the courts to the power of government and administrative decisions taken under its auspices. The earlier rejection of the idea of a body of administrative law has given way to an acceptance of the need for a set of rules which restrains the exercise of administrative power. In part this change can be traced to the influence of the European Convention on Human Rights and to the fact that the Treaty of Rome has become part of English law.

An English judge has declared that “we have now a developed set of distinct public law principles which are of general application, independent of private law and comparable to those in civil jurisdictions.”48 The High Court is now “exercising a quite separate jurisdiction: its inherent power to review administrative action.”49 He continued, “There are . . . situations where already, in upholding the rule of law, the courts have had to take a stand. The example that springs to mind is the Anisminic case [1969] 2 A.C. 147. In that case even the statement in an Act of Parliament that the Foreign Compensation Commission’s decision ‘shall not be called in question in any court of law’ did not succeed in excluding the jurisdiction of the Court. Since that case Parliament has not again mounted such a challenge to the reviewing power of the High Court. There has been, and I am confident there will continue to be, mutual respect for each other’s roles.”50 Another judge has said, “Thus, save as regards the Queen in Parliament, there is in principle always jurisdiction in the court to review the decisions of public bodies.”51

This assertion of judicial power has been felt in a number of ways. The courts have been ready to declare acts of ministers, usually in reality the actions of civil servants under the cover of formal ministerial powers, to be illegal, and not simply on the grounds that these acts are ultra vires but also that the courts exert a power of judicial review over administrative acts which “even the sovereign Parliament cannot abolish.”52

The Theory of Constitutionalism

Two main conclusions flow from the argument so far, conclusions which can be stated in a way which is valid both for Britain and the United States. First, there exists an entity, the “administration” (with a small a), which can be distinguished from the “policy branch.” In practice the administration already has a great deal of autonomy, is only marginally under the control of the policy branch, and could be more efficient if the existing, unsatisfactory link with “the government” were to be modified. Second, the present controls operated by the judiciary and the ombudsmen over the administration are inadequate to safeguard the rights of the individual. There needs, therefore, to be a new approach to the way government is articulated, and this in turn entails a quite different approach to the mechanisms by which control is exercised over the administration.

The traditional theory of the separation of powers sought to divide the functions of government between three branches of government and to keep the personnel of the three branches separate. The evident inability of this arrangement to control abuses by government led to the modification of the theory by grafting on to it checks and balances derived from the mixed constitution of eighteenth-century Britain. Although this institutional structure was explicitly embodied in the Constitution of the United States, the values it was intended to safeguard, democracy, efficiency, and justice, were just as important in nineteenth- and twentieth-century Britain, and the institutional structure of British government showed, and still shows, the influence of these values, particularly in the way in which particular processes characterize the operations of the differing branches of government. Although many commentators have rejected the idea that British government embodies a separation of powers, none would argue that laws should be made by civil servants, that members of the government should have the power to commit people to prison at will, or that the House of Commons or its committees should run the Health Service on a day-to-day basis. The development during the twentieth century of political parties that threatened the degree to which in reality the functions and branches of government were separated was not an overt attack upon the historic values they embodied, except by a small group of now discredited ideologues who rejected the whole basis of the system, but rather was a response to the introduction of a fourth value, social justice, which seemed to demand new structures and a new emphasis upon the co-ordination of the machinery of government, instead of the earlier emphasis on control. The consequence was the rise of the administrative state and the attack by its champions upon the separation of powers, an attack which as we have seen still characterizes those who yearn after an integrated, cohesive theory of administration.

Before beginning to develop the implications of these conclusions in detail, it would be as well to set out the general background approach to the view which I am adopting. First, the past fifty years have, to a considerable extent, been characterized by an optimistic, not to say utopian, view of human nature. On the Left, the assumption was that the abolition of property, and therefore the end of capitalism, would usher in a period in which, in the words of Karl Marx, communist society would make it possible for one to “do one thing today and another tomorrow, to hunt in the morning, fish in the afternoon, rear cattle in the evening, criticize after dinner, just as I have a mind, without ever becoming hunter, fisherman, shepherd or critic.”53 On the Right, extreme libertarians seem to believe that if government were abolished, the unrestrained free market would produce a universal harmony. Others seem to believe that if the administrators, the experts, were given free rein, without interference by politicians or judges, total efficiency would be achieved. There is no empirical evidence for any of these assertions, all of which derive presumably from the Victorian belief in the inevitability of the progress of man. In fact, the evidence suggests that human beings are characterized by ambition, by the desire for power, and by the search after ever greater satisfaction of wants. Human beings may be capable of improvement, but even if we make the assumption that they are perfectible, we are so far from realizing such a condition that to base political structures on the assumption of the perfectibility of man is just sheer foolishness. The world has not progressed very greatly since James Madison wrote, “in every political institution, a power to advance the public happiness involves a discretion which may be misapplied and abused.”54 Whether human nature is perfectible is an almost irrelevant consideration. We can agree that human nature is capable of being improved—Britain is a better society today than it was in the Middle Ages—but a glance at the circumstances of the world today is evidence enough that it would take a lot more than the abolition of property to produce utopia. Given this assumption of the frailty of human nature, it is surely wise to assume that those who hold power in government will be likely to abuse that power, not necessarily to the extent of a Hitler or a Stalin, although there are examples in the world today not far removed from those horrors, but in a thousand different ways.55

Second, institutions do make a difference. Institutions are well-established, rule-governed patterns of behaviour. If there were no institutionalized patterns of behaviour, there would be no predictability, all would be random, the true war of all against all. Institutions are not irrelevancies as Marxists and behaviourists would have us believe. It makes a great deal of difference to the individual accused of a crime whether he is tried in public, in court before a jury, rather than in secret by a member of the security police.

Third, if we need to guard constantly against the abuse of power, then we must heed the admonition of Montesquieu to oppose power with power. The recent history of British politics, with its revelations of corruption, mild perhaps in comparison with many other countries but there nevertheless in the legislature and in the government, confirms the continuing necessity for control mechanisms.

Fourth, it is important to emphasize that a commitment to constitutionalism does not consist in trying to increase the power of that particular part of the government machine or that political party or movement, which for the time being is following policies of which one happens to approve. The switch of conservative support to a strong presidency during the Reagan era is an example of this kind of misjudgement, based on the assumption that the presidency would remain under Republican control indefinitely. Thatcherite governments in Britain pursued policies which cast off traditional attitudes towards those constitutional conventions intended to limit the power of government.56 Thus they saw Friedrich Hayek simply as an economic guru but did not share his respect for constitutional principles.57 Constitutionalism is not a matter of seizing a short-term advantage; it is a belief in the need to establish and support those values in the political system which provide for stability and to maintain the procedures which protect the liberty of the individual in a democratic society. Thus it is important that opponents of the administrative state should not assume that it can be abolished. There is no way to predict how large, or how important, the administration will be in five or fifty years’ time: it may be larger, or it may be smaller, but it will be there. The priority, therefore, is to control it.

Equally, it should be accepted that the abuse of power can be perceived to arise from the actions of any branch of government—the policy branch, the legislature, the bureaucracy, or the judiciary. The exercise of presidential power to commit American troops abroad without congressional approval, from Theodore Roosevelt to Richard Nixon, resulted in passage of the War Powers Act of 1973. In recent years decisions of the United States Supreme Court have called forth many proposals to curb its power,58 and the recent challenges that the British judiciary has made to the power of both the policy branch and the bureaucracy have initiated charges of judicial usurpation.

It is equally important, however, to recognize the need for co-ordination of government as for control over government. The theory of the separation of powers and checks and balances emphasizes the element of control. The demand for co-ordination must be met largely through the mechanisms of politics—the political parties and pressure-groups—in other words through the interaction of the legislature and the policy branch.

The Control of the Administrative State

There have been two major developments in the nature of the political systems of America and Britain in the twentieth century. First, the structure of government has undergone fundamental change: administrative structures have developed to the extent that their impact upon the life of the ordinary citizen is more significant, and potentially more oppressive, than the actions of the traditional triad of governmental powers. The administration sits like a great cuckoo in the nest, elbowing out the historic actors in the drama of government. But this does not invalidate the analysis of the separation-of-powers theorist. It means that the analysis has to be brought up to date and applied to the new situation. The concern which always lay behind the doctrine of the separation of powers is still valid, namely, the concern to protect the individual against the overbearing power of government. Both models we have examined have failed to cope with the problem of the administrative state: in Britain the system of administrative government goes largely unchecked because of the pretence of ministerial responsibility; in the United States the separation of powers has allowed a greater degree of control over the administration, but “divided government” is a cause of deep concern to many political scientists, and it is asserted, the cause of public disquiet about the system of government.

Second, in the latter half of the twentieth century another fundamental change in the political systems of America and Britain has been taking place, a change which is gathering pace in the last decade of the century. There has been a decline in the importance of ideology in politics and consequently in the relevance of party organization to the conduct of government. It is true that there was a resurgence of ideology in the 1980’s at the elite levels of the party activists—Thatcherism and Reaganism dominated the headlines—but at the level of the electorate there has been a steady decline in the strength of party identification, and in the United States there has been the phenomenal development of split-ticket voting.

What is needed therefore is an analysis of government appropriate to an era in which the electorate is more concerned with outcomes than with party ideologies and government structures which reflect this concern. Students of administrative theory have bewailed the fact that “American public administration has been unable to develop a satisfying and enduring conception of democratic administration,”59 and have blamed this on the separation of powers, which built conflicting sets of values into the constitutional system, which pervade the administrative machine, and which could not be synthesized without “violating values deeply ingrained in the United States political culture.”60 It should not be assumed that the fact that these values, democracy, efficiency, and justice, are deeply ingrained is necessarily dysfunctional, even though they continually conflict. These values are there for good reasons, and they conflict because in practical situations it is rarely possible to hold that one or another of them is an absolute which must be maintained at all costs. Certainly one would not see administrative efficiency as an absolute which would override in all circumstances the considerations of justice. We have, therefore, to devise means, as best we can, which will reconcile these values within a viable institutional framework.

In the first half of the twentieth century administrative theorists argued for a system of government based on two functions—policy and administration.61 These writers were generally hostile to the traditional tripartite theory of the separation of powers.62 The politics-administration dichotomy does not work, because it is founded on an inadequate functional analysis, but some of the ideas these writers developed illuminate the possibilities of an administrative machine which would be relatively free of constant political interference but subject to external controls. It is the intention here, therefore, to attempt to synthesize these ideas into a more comprehensive analysis of the political system in the light of the nature of politics today.

As we have seen, there are four abstract functions of government and effectively four branches of government, although two of these, the policy branch and the administration, are still closely linked in terms of law, even if the reality is rather different. In the modern State we need a more effective control over the administrative machine, and the ending of the pretence of “accountability,” as we now understand it, would contribute towards this end. The autonomy of the administration should be recognized, but at the same time it should be subject to effective control. The administration needs to be independent to the extent that it is given a task to do and must carry it out. The policy branch should not be able to give direct instructions to the administration nor be responsible for the appointment, promotion, or dismissal of its members. What is being proposed here is a further extension of the process which was begun in Britain by the so-called Next Steps Report of 1988, produced by the Efficiency Unit of the Cabinet Office.63 The Report asserted that the responsibilities for management at the top of government departments were unclear; a few government ministers were even prepared to admit that they did not have the skills to manage their departments.64 Some top civil servants spent 90 percent of their time dealing with policy and “political support tasks” and only 10 percent managing their organizations. The Report recommended that agencies should be established “to carry out the executive functions of government within a policy and resources framework set by a department.”65 The Report made an attempt to deal with the question of the accountability of these agencies to Parliament, but it could not break away from the accepted theory of the Constitution: “It is axiomatic that Ministers should remain fully and clearly accountable for policy. For agencies which are government departments or parts of departments ultimate accountability for operations must also rest with Ministers.”66 Nevertheless the authors of the Report hoped that a convention could be established so that heads of agencies could have delegated authority from Ministers for the operation of their agencies. “There is nothing new in the suggestion that Ministers should not be held answerable for many day-to-day decisions involving the public and public services.” This deliberate blurring of the issue of accountability in order to pay lip-service to the doctrine of ministerial responsibility, typical of much of the way in which the unwritten British “Constitution” now operates, was further compounded by the Prime Minister, Margaret Thatcher, when she introduced the Report to the House of Commons. She assured the House that “there will be no change in the arrangements for accountability. Ministers will continue to account to Parliament for all the work of their Departments, including the work of the agencies.”67 The Efficiency Unit, however, which was attempting to drive forward the agency concept, returned to the task of reducing the interference of ministers in the administrative process. The terms of reference of a study set up in 1990 to assess the progress of the “Next Steps” initiative proclaimed that “Ministers and Departments should to the greatest extent practicable stand back and leave Agency managers free to manage. Intervention, planned or unplanned, in the day-to-day managements of Agencies should be exceptional and positively justified in each case.”68 The study group concluded that “Chief Executives are directly accountable to a Minister in a ‘quasi-contractual’ relationship. In any such relationship it is important that the responsibilities of both sides are clearly defined and both are in a position to deliver their side of the bargain.” Thus a typical British compromise was evolved, blurring the issues at stake. The minister remained accountable to Parliament, but it was hoped that the minister would behave reasonably and not interfere in the day-to-day administration of agencies.69 This hope was shattered in October 1995 when the Home Secretary dismissed the Director General of the Prison Services Agency for his failure to administer the Prison Service efficiently. This dismissal occurred amidst accusations that the minister had intervened in the day-to-day operations of the Agency, without being prepared to accept responsibility for the failure of its actions.

Thus we have arrived at a point where either the government must accept full responsibility for every action of the administration, however trivial, which is unrealistic, or it must be detached altogether from its operation. The middle ground is unacceptable and unworkable. True we must accept that it would not be possible for the administration to carry out its tasks without exercising judgement and discretion and that “politics” would characterize its operation just as it does in any large organization. For that reason it is vital that the administration should be subject to effective checks. The policy branch would be required to report to the legislature about how the administration was working and to propose reforms or policy changes. The administration would report to the legislature annually and would be subject to all the normal procedures for control of finance, and the committees of the legislature would scrutinize its operation. This process of committee scrutiny could certainly be much more effective in Britain than it is at present, because committee investigations would no longer be seen as direct criticism of ministers. The policy branch would make recommendations to the legislature both for changes in policy and in administrative procedures, and the administration would be subject to review by the courts, as suggested below.

The sphere of the “administration” would exclude those activities of government where what I have called the discretionary function is dominant. It would be impossible to include all the activities of government within the framework here proposed. Foreign affairs, defence, and the macroeconomic control of the economy necessitate government action which cannot be subject to the restraints which are necessary in the fields of domestic regulation and social policy, and these activities would remain under the direct control of the policy branch.

There are two problems with this approach. If policy-makers are divorced from those who will have to put the policies into effect the results will be unrealistic and uninformed by the requirements of implementation, and when these policies result in failure, the electorate will not know whom to blame. But accountability is already unsatisfactory in both the United States and Britain, in spite of the differing extent to which the separation of powers is embodied in their respective institutional structures. In the United States the President blames Congress for failing to legislate his programmes, and Congress blames the President for failing to implement legislation effectively. In Britain, as we have seen, the close alliance of government ministers and government members of the House of Commons makes effective control impossible. Thus what is required is a clear and open procedure for taking policy decisions so that responsibility can be unambiguously allocated. For this reason it is essential that the “policy branch” consults the “administration” before proposing legislation, obtains its views in writing, and is required to publish them with all draft legislation. It would be a mark of the existence of a mature society that policy decisions, other than those affecting foreign affairs and defence, should be taken only after public scrutiny of all the considerations involved, including the opinions of those administrators who would be charged with the implementation of the policy and who would have the independence that would evoke an honest opinion.

The second problem lies in the apparent assumption that the administrative machine does not have policy aims of its own and will faithfully implement the policies laid down by the legislature. Thus although the checks to the abuse of power by the administration set out above are important at a general level, if individual rights are to be safeguarded there must be a much more detailed case-by-case control. One solution to the problem would be to follow the continental system of administrative courts, the model of the French Conseil d’Etat, which is quite separate from the general courts.70 The danger, however, would be that such a court would seem totally alien to the Common Law tradition. Another solution is possible. Both the British and American courts have shown a readiness to develop the kind of public law which could fill the gap left in our system of jurisprudence. At present administrative cases come piecemeal to the United States Federal District Court, or to the English High Court, are subject to all the corresponding problems of delay and expense and are dealt with by judges who are not specialists in administrative law. Yet our systems of courts already include specialist divisions, which because of their differing subject material have differing procedures. It would be perfectly possible to set up an Administrative Division with the tasks of overseeing the administration’s activities, hearing complaints from the public, and providing remedies. It ought not to operate through the adversary system, which characterizes our current judicial system, and it could have very different methods of gathering evidence, accepting documents and written submissions where possible. The Division could in fact learn a great deal from the procedures of the Conseil d’Etat but still be integrated into the body of the judicial system. Appeals could lie to the Common Law courts, but the grounds of appeal, particularly by the administration, could be severely limited to important questions of law.

It is not possible to allocate particular functions exclusively to each branch of government (and in the United States this is not merely because the Constitution makes specific exceptions to the overall distribution of power), but it is possible to say that there is a function which is more appropriate to a particular procedure, to attempt to restrict each branch to particular procedures, and therefore to make one function the dominant concern of that branch. In a sense this is the aim of those who have espoused the “functionalist” tendency on the Supreme Court in recent years, but they have attempted to achieve it without setting out a clear basis for their decisions.

Thus we can accept that the rule-making function is exercised in some degree by all branches of government but nevertheless assert that the legislature should be concerned only with rule-making of a general kind and that the rules it makes should be binding on both the policy branch and the administration and subject to being over-ruled by the judiciary only on the grounds of their having offended against certain basic constitutional principles. That the legislative branch should restrict its rule-making to general rules, not dealing with specific individuals, is an essential part of the Rule of Law, a major tenet of constitutionalism for 350 years. Thus those legislative vetoes of the United States Congress which, as in the Chadha Case, dealt with specific individuals should have been invalidated because they dealt with particular cases. The Supreme Court invalidated the legislative veto on the grounds that one part of the legislature was exercising a “legislative function” which should have been exercised only by both Houses and with the approval of the President. On the contrary, what the House of Representatives did in exercising a “veto” over the granting of resident status to Chadha was to engage in rule-application.

Justice Powell in his concurring opinion in Chadha made this point clearly, although he thought the use of the veto in this case constituted rule-adjudication. “On its face, the House’s action appears clearly judicatory. The House did not enact a general rule. . . . It thus undertook the type of decision that traditionally has been left to other branches.”71 He quoted John Marshall in Fletcher v. Peck, “It is the peculiar province of the legislature to prescribe general rules for the government of society; the application of those rules to individuals in society would seem to be the duty of other departments.”72 Powell made reference to the constitutional prohibition on Bills of Attainder and quoted United States v. Brown, to the effect that the separation of powers was intended as a safeguard against “trial by legislature.”73 In other words this was an inappropriate procedure for making such a decision. This surely is the basis of the Rule of Law. It is the consideration which led to the rejection of the exercise of the unlimited power of the Long Parliament and of the American revolutionary state legislatures. There can be no Rule of Law if the legislature intervenes arbitrarily in the administrative or judicial process to favour or disadvantage an individual, as in the Chadha Case.74

Conclusion

Compared with the parliamentary system the weakness of the American version of the separation of powers lies in its apparent inability to deliver a satisfactory degree of co-ordination between the legislative branch and the policy branch, even to the extent that government can be on the edge of collapse, as in 1995. The American system does, however, provide the possibility of an effective control over the policy branch and the administration, which the Westminster model of parliamentary government has signally failed to achieve. The problem at the centre of constitutional government today, as has been for the past 350 years, is how to achieve a balance between co-ordination and control in the relationships among the branches of government which will safeguard individual freedom but which will also ensure that government can deliver to its citizens those essential services without which modern society cannot survive.

A major function of the legislative branch is to exercise control over the policy branch and the administration. The effectiveness of this control is hampered by the partisanship of members of the legislature, a partisanship which is artificial, not reflecting the needs or attitudes of the electorate at large. Separating the policy branch from the main body of the administrative machine would tend to diminish the effect of partisanship in the legislature’s attitudes towards the administration. This arrangement would decrease very considerably the size of the overblown ranks of the present governments in Britain, where almost one-third of the government members of the House of Commons may hold an office of profit under the Crown. The growth in the number of members of the Commons in the government has, of course, been a consequence of the growth of the administrative state, and inevitably the need to find such a large proportion of the governing party in the Commons to run the administration has led to second-rate people being put into posts for which they do not have the necessary competence. To return many of these people to the back benches would reduce the government’s power of patronage and would therefore be likely to increase the independence of government back-bench members. The problems faced by an incoming President, when control of the presidency changes from one party to another, involving the filling of thousands of positions, would also be alleviated by separating the administration off, consequently making more posts into civil-service positions.

To hive off a major part of the administration from the direct control of political leaders would transform the character of the British “Government” and the American “Administration.” It would remove from them the day-to-day routines of the administrative machine, which they are in general ill-equipped to supervise, and allow them to concentrate upon those vital functions which only the “policy branch” can perform: the planning of legislative policy and the formulation and implementation of policy concerning foreign affairs, defence, and macroeconomics.

Another major responsibility of the policy branch is co-ordination, working through the party system and pressure-groups. The problem of co-ordination may be seen at two levels—the over-all co-ordination of legislative and administrative policy, for example, through the budget, or the co-ordination of the activities of the different departments of the administration. The problem of co-ordinating the policies of the presidency and the Congress has in recent years concentrated attention on the phenomenon of divided government and on the desirability, in the eyes of some commentators, of strengthening the party system in order to prevent this from occurring. Divided government may be the result of purposive voting or an “accidental” result of the separation of powers in a period of weak party government.75 “Purposeful voting advocates put their faith in the will of the majority: as long as the people want (divided government), that’s fine,”76 but in the opinion of its critics the problems of the American system are not always the product of divided government: “The separation of powers operating in an age of weak parties is as likely a culprit.”77 But if this is the case it is important to ask why the political parties are so weak. The root cause of the weakness of American political parties is the increasing fragmentation of the electorate. America has experienced increasing diversity in the electorate, of a number of different kinds, since the Second World War. Ethnic groups previously excluded from the political process are now important electoral forces. The multi-cultural society has replaced the old system of positive Americanization, which from the time of Jefferson has been the all-embracing ideology of the United States. It was not until the Vietnam experience that it became impossible to impose this positive Americanization through the processes of socialization that had previously operated. The resulting fragmentation has made coalition government inevitable, and it is unlikely that such deeply embedded social factors can be overcome by tinkering with the electoral system. Coalition government everywhere reflects potentially irreconcilable conflicts in society; why should it be different in the United States? A parliamentary system with weak parties would be no better. The great benefit that the Constitution has given to the United States is stability, and this becomes even more important in an era when the electorate is increasingly fragmented.

Unless divided government can be eliminated, which seems very unlikely, then the less coherent the parties are, the better. In the long run President and Congress must live together by compromise, and the more coherent and ideological the parties become, the more difficult this will be. The “great” periods of united control, such as the New Deal period, actually worked only because party unity was not complete, allowing the President to find support where he could. To adopt the system of a quadruple separation of powers advocated here should make it easier to live with a fragmented political system and at the same time keep the powers of government from abusing their position.

There are two dangers in giving the administration a greater degree of autonomy. First, the danger that pressure-groups would turn their attention to the administrative machine to an even greater extent than at present.78 The second is that the “directors-general” of the administrative departments, meeting together to co-ordinate their operations, would become the unelected government of the country. These dangers could be offset if the legislature, freed from the sense of partisanship in relation to the administration, exercised a genuine control over that administration and would be able to work more closely and effectively with the President. Equally, the convergence of the political parties that is taking place in Britain in the second half of the 1990’s offers the possibility of a system of government in which genuine accountability to the legislature, and ultimately to the electorate, may make the control of government more of a reality than has been the case since the rise of the administrative state.

[1. ]Peter Winch, The Idea of a Social Science and Its Relation to Philosophy, London, 1958.

[2. ]Quentin Skinner, “Meaning and Understanding in the History of Ideas,” History and Theory, Vol. VIII, No. 1, 1969. See also John Dunn, review of The Meaning of the Separation of Powers, by William B. Gwyn, The Historical Journal, 1967, pp. 472–4.

[3. ]Skinner, op. cit., p. 48.

[4. ]Ibid., p. 52.

[5. ]Ibid., p. 50.

[6. ]Quentin Skinner, The Foundations of Modern Political Thought, Vol. 1, The Renaissance, Cambridge, 1978, p. xi.

[7. ]See the section of the Bibliography on Modern American Constitutional Law and Theory, pp. 436–40 below.

[8. ]Philip B. Kurland, Watergate and the Constitution, Chicago, 1978, p. 34.

[9. ]United States v. Nixon, 418 U.S. 683 (1974), at p. 708.

[10. ]Ibid., p. 713. See also Nixon v. Administrator of General Services, 433 U.S. 425 (1977).

[11. ]K. A. Kirwan, “The Use and Abuse of Power: The Supreme Court and Separation of Powers,” The Annals of the American Academy, Vol. 537, 1995, p. 78.

[12. ]Immigration and Naturalization Service v. Chadha, 462 U.S. 919 (1983).

[13. ]Ibid., pp. 957–8.

[14. ]Louis Fisher, “The Legislative Veto: Invalidated It Survives,” Law and Contemporary Problems, Vol. 56, No. 4, autumn 1993.

[15. ]Bowsher v. Synar, 478 U.S. 714 (1986), at p. 722.

[16. ]Ibid., p. 734.

[17. ]K. J. Harriger, “Separation of Powers and the Politics of Independent Counsels,” Political Science Quarterly, Vol. 109, No. 2, summer 1994.

[18. ]Morrison v. Olson, 108 S.Ct. 2597 (1988). See below, p. 402.

[19. ]Allen v. Wright, 468 U.S. 737 (1984).

[20. ]Ibid., at p. 761.

[21. ]Cass R. Sunstein, “Standing and the Privatization of Public Law,” Columbia Law Review, Vol. 88, No. 6, October 1988, pp. 1460–1. See also C. J. Sprigman, “Standing on Firmer Ground—Separation of Powers and Deference to Congressional Findings in the Standings Analysis,” University of Chicago Law Review, Vol. 59, No. 4, 1992.

[22. ]Sunstein, op. cit., p. 1461.

[23. ]Mistretta v. U.S., 109 S.Ct. 647 (1989).

[24. ]Public Citizen v. U.S. Department of Justice, 491 U.S. 440 (1989).

[25. ]Association of American Physicians and Surgeons v. Clinton, 997 F.2d 898 (D.C. Cir. 1993). See J. S. Bybee, “Advising the President: Separation of Powers and the Federal Advisory Committee Act,” Yale Law Journal, Vol. 104, No. 1, Oct. 1994.

[26. ]Filartiga v. Peña-Irala, 630 F.2d 876 (2d Cir. 1980). See H. H. Koh, “Transnational Public Law Litigation,” Yale Law Journal, Vol. 100, No. 8, 1991, pp. 2362–8.

[27. ]Ahmad v. Wigen, 726 F. Supp. 389 (EDNY 1989).

[28. ]Plaut v. Spendthrift Farm, Inc., 115 S.Ct. 1447 (1995). See also Franklin v. Massachusetts, 112 S.Ct. 2767 (1992); Touby v. United States, 500 U.S. 160 (1991); and Metropolitan Washington Airports Authority v. Citizens for the Abatement of Aircraft Noise, Inc. (1991).

[29. ]For a very different view of the “balance” between the branches of government see Ralph Rossum, Congressional Control of the Judiciary: The Article III Option, Center for Judicial Studies, Cumberland, Va., 1988.

[30. ]David McKay, “Review Article: Divided and Governed? Recent Research on Divided Government in the United States,” British Journal of Political Science, Vol. 24, No. 4, 1994.

[31. ]James Q. Wilson, “Does the Separation of Powers Still Work?” The Public Interest, Vol. 86, winter 1987, p. 43.

[32. ]Lloyd N. Cutler, “To Form a Government,” Foreign Affairs, fall 1980, pp. 126–7.

[33. ]See M. J. C. Vile, “Presidential and Parliamentary Systems,” in Albert Lepawsky (ed.), The Prospect for Presidential-Congressional Government, Berkeley, 1977.

[34. ]William B. Gwyn, “The Separation of Powers and Modern Forms of Democratic Government,” in Robert A. Goldwin and Art Kaufman (eds.), Separation of Powers—Does It Still Work? Washington, D.C., 1986.

[35. ]Since 1895 no government with an over-all majority in the House of Commons has been forced to resign by losing a vote in the House.

[36. ]The House of Lords can exercise a check upon governments, and in a number of cases caused the government to change its policy, but the extent to which governments accept changes in policy rather than over-ruling the Lords depends in the last resort on the internal politics of the government party in the Commons.

[37. ]This attitude is perfectly compatible with the (very flexible) conventions of the British Constitution. “What the Government will treat as a matter of sufficient importance to demand resignation or dissolution is, primarily, a question for the Government.” Sir Ivor Jennings, Cabinet Government, Cambridge, 1947, p. 381.

[38. ]Similar palace revolutions occurred in 1976 when James Callaghan succeeded Harold Wilson and in 1990 when John Major replaced Margaret Thatcher. Formerly appointments of new Prime Ministers had taken place without reference to Parliament or without a general election, but that was before the Conservative Party moved to elections for its Leader and in an age when the role of the monarch was still seen as significant in the choice of Prime Minister.

[39. ]Select Committee on the Public Service, House of Commons, Ministerial Accountability and Responsibility, Second Report, 1996, p. xvi.

[40. ]Mark Freedland, “Privatising Carltona: Part II of the Deregulation and Contracting Out Act 1994,” Public Law, spring 1995, p. 21.

[41. ]See pp. 304–8 above.

[42. ]Gary Lawson, “The Rise and Rise of the Administrative State,” Harvard Law Review, Vol. 107, 1994, p. 1253.

[43. ]William B. Gwyn, “The Indeterminacy of the Separation of Powers and the Federal Courts,” George Washington Law Review, Vol. 57, No. 3, 1989, pp. 474–5.

[44. ]Dissenting opinion in Morrison v. Olson, 108 S.Ct. 2597 (1988) at p. 2627.

[45. ]Ibid., p. 2641.

[46. ]See above pp. 358–9.

[47. ]Jesse H. Choper, Judicial Review and the National Political Process: A Functional Reconsideration of the Role of the Supreme Court, Chicago, 1980, p. 263. See also Gwyn, op. cit., pp. 504–5. For an extended discussion of this aspect of the problem see Thomas W. Merrill, “The Constitutional Principle of the Separation of Powers,” The Supreme Court Review, 1991, Chicago, 1992, especially pages 226–9.

[48. ]Lord Woolf of Barnes, “Droit Public—English Style,” Public Law, spring 1995, p. 57.

[49. ]Ibid.

[50. ]Ibid., p. 69.

[51. ]Sir John Laws, “Law and Democracy,” Public Law, spring 1995, p. 72.

[52. ]Sir William Wade and Christopher Forsyth, Administrative Law, 7th edn., Oxford, 1994, p. 737. See also Richard Gordon, “The New Sovereigns?” New Law Journal, Vol. 145, Apr. 14, 1995, p. 529.

[53. ]The German Ideology, quoted in Erich Fromm, Marx’s Concept of Man, New York, 1961, p. 42.

[54. ]Federalist, No. 41.

[55. ]Any theory of politics must begin with a discussion of human nature. The “public choice” school of political science has entered into the discussion of the separation of powers. See G. Brennan and A. Hamlin, “A Revisionist View of the Separation of Powers,” Journal of Theoretical Politics, Vol. 6, No. 3, July 1994. The assumptions on which they base their analysis are so far removed from reality as to make their conclusions of limited value. An attempt to overcome these limitations is the development of models based on more complex assumptions. See M. C. Jensen and W. H. Meckling, “The Nature of Man,” Journal of Applied Corporate Finance, Vol. 7, No. 2, 1994.

[56. ]See Cosmo Graham and Tony Prosser (eds.), Waiving the Rules: The Constitution Under Thatcherism, Milton Keynes, 1988.

[57. ]F. A. Hayek, Law, Legislation and Liberty, London, 1989. See also Graham and Prosser, op. cit., pp. 17–18; and Richard Bellamy, “‘Dethroning Politics’: Liberalism, Constitutionalism and Democracy in the Thought of F. A. Hayek,” British Journal of Political Science, Vol. 24, No. 4, 1994.

[58. ]See Ralph Rossum, Congressional Control of the Judiciary: The Article III Option, Cumberland, Va., 1988.

[59. ]Robert S. Kravchuk, “Liberalism and the American Administrative State,” Public Administration Review, Vol. 52, No. 4, July/Aug. 1992, p. 374.

[60. ]David H. Rosenbloom, “Public Administrative Theory and the Separation of Powers,” Public Administration Review, Vol. 43, May/June 1983, p. 219.

[61. ]See pp. 304–8 above.

[62. ]See Laurence J. O’Toole, “Doctrines and Developments: Separation of Powers, the Politics-Administration Dichotomy, and the Rise of the Administrative State,” Public Administration Review, Vol. 47, Jan./Feb. 1987, pp. 17–19.

[63. ]Improving Management in Government: The Next Steps. Report to the Prime Minister, The Efficiency Unit, HMSO, London, 1988. See E. C. S. Wade and A. W. Bradley, Constitutional and Administrative Law, 11th edn., London, 1993, pp. 294–5.

[64. ]Improving Management in Government, op. cit., p. 25.

[65. ]Ibid., p. 9.

[66. ]Ibid., p. 17.

[67. ]Hansard, 18 Feb. 1988, Col. 1151.

[68. ]Making the Most of Next Steps: The Management of Ministers’ Departments and Their Executive Agencies, Report to the Prime Minister, The Efficiency Unit, HMSO, 1991, p. 30.

[69. ]Ibid., p. 7.

[70. ]See pp. 258–9 above.

[71. ]Chadha, op. cit., at pp. 964–5.

[72. ]Fletcher v. Peck, 6 Cranch 87 (1810), p. 136.

[73. ]Chadha, op. cit., at p. 962.

[74. ]The importance of art. 1, sec. 9, of the Constitution, the Bill of Attainder prohibition, was fully discussed in Nixon v. Administrator of General Services. The Court distinguished the legislation under consideration, The Presidential Recordings and Materials Act of 1974, from a Bill of Attainder, arguing that the Act did not inflict a “legislative punishment” on Mr. Nixon. 97 S.Ct. 2777 (1977), pp. 2803–11.

[75. ]James L. Sundquist, “Needed: A Political Theory for the New Era of Coalition Government in the United States,” Political Science Quarterly, Vol. 103, 1988, p. 527.

[76. ]McKay, op. cit., p. 533.

[77. ]Ibid., p. 534.