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

TWELVE

A Model of a Theory of Constitutionalism

Possibly the dominant impression left by this survey of three centuries of Western constitutional thought is that we of the middle of the twentieth century live in an age which has inherited a number of different traditions, without being able ourselves to knit these varied strands into a coherent pattern, to derive a unified reconciling theory of constitutional government. Some prevailing intellectual modes, indeed, are hostile to any attempt at it. The demands made upon modern governments, the complexity of their aims, demand new techniques, new procedures, new forms: it is not surprising that the relevance of older theories and traditional systems of thought are doubted, and certain modern students of government have adopted a narrowly conceived “scientific” approach to the study of politics which is intended as a fresh start. Yet their own work, as we have tried to show, betrays the impossibility of rejecting centuries of discussion on a subject as purposive as the pursuit of political aims. Their value-patterns and aspirations show through the superficial detachment, and they reveal themselves as the children of Locke, of Montesquieu, and of Madison.

For the functional categories of the doctrine of the separation of powers with their intimate relationship to the rule of law, the concept of balance which was the essential element of theories of limited government, and the central ideas of representation and responsibility underlying theories of parliamentary government—all these continue to be important parts of our intellectual apparatus. We still appreciate the ideal of a moderate government, one which will avoid the extremes of any “simple” form of State—an ideal to which the ancient world first gave expression. The demand for freedom from arbitrary rule, which dominated the minds of the men of the post-Renaissance era, is also our demand. The nineteenth-century middle-class aspiration towards a “harmonious” polity is, in the conditions of universal suffrage, transmuted into an increasingly insistent search for co-ordinated and efficient government.

The persistence of these values—justice, “democracy,” efficiency—in our constitutional thinking does not, of course, demonstrate that the old functional and structural conceptions, which earlier writers found satisfactory, are still adequate to explain modern government. New trends in the institutional development of advanced Western countries in the last hundred years are somewhat difficult to fit into the older categories, and new concepts have to be found. Account must be taken of the concept of process, which has so illumined modern sociological thought, and which can now be put alongside the older concepts of function and structure.

These, then, are the three elements of the model which it is proposed to develop: to approach the constitutional system from these three related view-points of function, structure, and process; to show how they are interdependent, mutually interacting, and how they are intimately related to certain value-patterns; and to emphasize that the character of constitutionality lies in this inter-penetration of function, structure, and process.

In the history of constitutional theory the most persistently used concept has been that of function. It finds its roots in Greek political thought, it is the basis of the idea of a government of laws, and it has been the most used tool of analysis for purposes of articulating the parts of government. Yet it has been subjected to a vast amount of criticism, and by many writers has been rejected as a useless concept. Perhaps the most serious aspect of these criticisms is that they strike at the basis of the rule of law, not merely in the sense of the particular formulation that Dicey developed, but in the sense in which a primary rule-making function is the basis of any constitutional theory.

The nature of the “functions” of government thus requires considerable clarification. The long discussion of the powers of government has been conducted largely in terms of the legislative, executive, and judicial functions. These abstract concepts emerged after a long period in which men thought mainly in terms of the “tasks” which government had to perform, such as conducting war and diplomacy, and maintaining order. The emergence of the idea of legislative and executive powers, or functions, had in itself little to do with an analysis of the essential nature of government; it was concerned more with the desire, by delimiting certain functional areas, to be able to restrict the ruler to a particular aspect of government and so to exercise limits on his power. This “purposive” quality of the traditional classification of government is important, for it makes the discussion of functional analysis much more than simply an attempt at description; it inevitably carries a normative connotation as well. The very use of these terms assumes a commitment to some form of constitutional government.

Let us, for a moment, accept the traditional triad of “government functions” in order to consider the problems of adopting a functional analysis. However, we shall think in terms of the functions of the political system rather than the government, for the problem of the control of government action which lies at the heart of constitutional thought necessitates an over-all view of function, rather than a concern merely with the relationship between government and citizens implied in the earlier usage. This emphasis upon the political system, rather than upon the machinery of government alone, is a characteristic of a recent functional analysis of government, that of Gabriel Almond in The Politics of the Developing Areas. Almond attempts to incorporate into the functional concepts developed by political theorists the rather different type of functional analysis of the sociologist. He distinguishes the “input” functions of the political system from the governmental “output” functions. The former include political socialization and recruitment, interest articulation and aggregation, and political communication. The latter consist of the rule-making, rule-application, and rule-adjudication functions;1 Almond attempts to rid these concepts of their attachment to particular structures of government by adopting these terms instead of the more familiar legislative, executive, and judicial categories. This usage helps to make clear the often-emphasized fact that the structures of government are, in Almond’s term, “multi-functional.” That is to say that rules are made by civil servants and by judges as well as by legislatures; rules are applied by the courts as well as by “the executive”; and judgements are made by civil servants and ministers as well as by judges. Thus the purposive nature of the older concepts is removed and the scientific generality of these tools for the analysis of systems of government is established. Yet, as we have seen, the values of earlier centuries persist strongly, and the insistence upon rules being made by legislatures remains for Almond and Coleman a distinguishing characteristic of democratic political systems. The fact that a particular task of government is regulated by “legislation” rather than by some other procedure reflects the determination that certain values shall predominate in the ordering of society rather than others. It is in fact the procedures of government which are “purposive,” for they are chosen or rejected to perform certain tasks because of the values they embody. It is this connection between functions, procedures, and values which we must explore.

The pure doctrine of the separation of powers implied that the functions of government could be uniquely divided up between the branches of government in such a way that no branch need ever exercise the function of another. In practice such a division of function has never been achieved, nor indeed is it desirable that it should be, for it would involve a disjuncture in the actions of government which would be intolerable. But the criticism of the threefold conception of government functions can be taken much further than pointing out that it has never been wholly achieved in practice. It can be suggested that the “multifunctionality” of political structures can, and perhaps must, be carried to the point where any attempt at a division of functions is quite impossible.

Let us take two examples, the first from the courts. A judge when dealing with a case, at any rate in a common-law country, is applying the law to a particular instance, is “judging” or determining the nature of the rule to be applied, and is at the same time creating a precedent to be followed by other courts. He, therefore, of necessity, exercises all three functions, and cannot be prevented from doing so if he is to perform the tasks which he is set. To take a more extreme case, let us consider the work of the administrator. The most extreme theorists of the policy-administration dichotomy suggest that the civil servant merely exercises a technical skill which is directed towards the execution of rules laid down for him by the political branches of the government. Thus they think in terms of an “administrative function.” But in fact civil servants, without any intention of abusing their powers, inevitably make rules, interpret them, and apply them. Nor is this limited to the formal cases of delegated legislation, or administrative jurisdiction, which have received so much attention. As in the case of the judge, it seems to be inherent in the tasks which the administrator must perform. The administration of any complex governmental programme involves the taking of many decisions at all stages of its operation, many of which will become precedents for later administrative decisions. These “rules,” though not always formulated as such will govern the decisions of other civil servants, and will be applied to the clients of the government department, the public, who are affected by its decisions. This situation arises from the fact that statutes or other policy decisions can never present a rigid plan to be followed exactly by civil servants, who must be allowed a certain discretion, often very wide, in the administration of government programmes. Furthermore, administrative action demands a high degree of consistent, patterned behaviour, and the administrative machine generates the rules and precedents which will ensure this consistency. They will not be “authoritative” in the sense in which a court ruling is authoritative, for they could in principle be invalidated by judicial review. Yet as far as the ordinary citizen is concerned they will for the most part represent “the law.” Often these rules will be matters on which a court would refuse to adjudicate, and the overwhelming majority will never be brought before the courts.

The misconception of the functional categories of the separation of powers, therefore, stemmed from the naive view that there were distinct actions of willing and execution that could be isolated and kept in separate compartments. Such a distinction might possibly be applied to extremely simple actions. A decision to open the door, and then opening it, might be treated in this way. But most operations of government are much too complex, requiring a whole stream of decisions to be taken, such that it is impossible to divide them up into acts of will and acts of execution. In the same way the later distinction between “policy” and “administration,” which was intended to replace the legislative-executive dichotomy, also breaks down. An important decision about “policy” will often be the climax of much administrative activity, rather than the initiation of it.2

Thus if we pursue the analysis to its limits we see that the exclusive allocation of rule-making, rule-application, or rule-adjudication to particular organs of government is not only inconvenient, it is probably quite impossible. Every act of every official, except perhaps the most routine and trivial operations, embodies all three types of activity. The policeman on the beat creates precedents in his actions, even if only for the people in his vicinity; he determines in which cases he will apply the rules, and when he does apply them his decision is subject to appeal; but when he decides not to apply the rules, as he interprets them, his interpretation will usually be final. Thus within his sphere of competence he will make rules by interpretation of other rules, and apply them. Such a view might seem to put an end to functional analysis of the kind which has in the past been the basis of attempts to confine parts of the government to specified types of behaviour.

Can we then save the idea of functionalism in the sense in which it has traditionally been used in the analysis of political systems? The first idea which we might use is that of a hierarchy or structure of rules, so that even if we accept that judges and administrators must also make rules these will be subordinated to those made by the legislature. This is, of course, the extension of the idea of law which has for centuries been the sheet-anchor of the concept of constitutionalism; the idea of a hierarchy of norms that will enable each of the decisions of an official to be tested against a higher rule. This ultimate rule-making authority in a democratic system is entrusted to a body representative of the people, or to a constitution ultimately subject to popular control. The history of Western constitutionalism has been the history of the constant pressure to maintain the ultimate authority of “the legislature.”

However, there are considerable difficulties with this seemingly essential element of the theory of the constitutional State. To argue that rules must be arranged in a hierarchical fashion so that they can be tested suggests an ability always to apply the logical criteria of deduction to the material of government decisions. It can easily be shown from the material of judicial opinions that the discretion in the hands of a judge to place one interpretation rather than another upon the relationship between “higher” and “subordinate” decisions is often very great indeed. Yet the very fabric of the constitutional State would seem to depend upon the willingness of courts to undertake and to operate successfully this “semi-logical” procedure. The success of the Supreme Court of the United States in exercising this discretion has been considerable, and in spite of the highly-charged political character of many of its decisions it has retained the ability and the prestige to enable it to perform this function. In the field of administrative law, in which so many of these decisions fall, the Conseil d’Etat in France has been equally successful. In Britain, however, one of the most marked governmental trends has been the continued decline of judicial power—a refusal on the part of the judges to exercise this discretion and a readiness to acquiesce in governmental and administrative acts. Discredited in their attempts at defending privilege against government action, the courts have not followed the line the American courts took in the same situation, and become the defenders of “the individual,” or “the people,” against a government which might be representing interests rather than the community. Part of this failure has undoubtedly been a failure of nerve, but partly also it is due to the operation of party politics, which has increasingly threatened the judges with the use of the rule-making power of Parliament if they attempted to restrain administrative actions. There have been a number of instances since the Second World War where decisions of the courts which went against the administration have been quickly reversed by Parliament, sometimes with incredible rapidity. The case of the War Damage Act of 1965 in Britain is an extreme one, for it had retroactive effect, and so offended against the basic rule of constitutionalism, the predictability of action.

Thus even if we accept “the rule of law” in the sense of a hierarchy of rules which can ultimately be tested against the final statutory or constitutional authority, we have to face the fact that this may become merely a matter of form. The outward forms of legality are retained, but they may be manipulated by the party leaders in a way which makes a mockery of them. The twentieth-century concern with the solution of practical problems can be taken to the point where the desire to deal with particular issues overrides all consideration of the way in which these issues are dealt with. Yet so long as the procedures for changing the law, or for making new laws, retain their vitality, and represent a genuine check upon the bureaucracy and its political chiefs, the dangers of the abuse of power in this way are minimized.

Thus, although dangers exist, it would be wrong to argue that at the present time all powers in Britain are accumulated in the same hands. There are two reasons for taking this view. The first lies in the nature of the ultimate rule-making power. It is obvious that neither in France, in the United States, nor in Britain does “the legislature” any longer exclusively exercise a rule-making, or legislative power. In Britain, of course, the “King-in-Parliament” legislates in the formal sense, but the House of Commons does not, either alone or in combination with the “upper” house, actively make law. “Legislative supremacy” is guarded, but the “legislature,” in the sense of the assembled representatives of the people, does not write the laws. The overwhelming proportion of legislation is written in the government departments and is presented to Parliament by ministers and accepted by the House with only minor amendments. In the United States the trend towards Presidential legislation is also marked, although it has not, of course, developed to the same extent that it has in Britain. Most major measures are prepared by the Administration, and although a particular session of Congress may exercise its prerogatives and substantially amend or even reject Administration proposals, it seems to be true that constant pressure by Administrations over a number of years will gradually wear down Congressional opposition. We are perhaps entering a period when the most significant difference between the British and American systems of government is that in Britain consultation about proposed legislation takes place almost exclusively before its introduction into Parliament, whereas in America it still takes place within Congress. But the ultimate result is still that the major initiative for, and the largest share in the framing of, legislation is in the hands of civil servants, ministers, and presidents or their advisors. In Fifth Republic France the need for rules to be made other than in the “legislature” has been recognized by the granting of a constitutional rule-making power to the government.

But have we really described how rules are made in Britain? What does rule-making mean? Are rules made by the person who drafts them, by the body which formally approves them, or by the leaders who instruct the draftsmen and who organize the approving body? The answer in Britain today is, surely, that the rules are made by all three elements in the process. We should not forget that ministers and civil servants are different bodies of men, differently composed, differently recruited, with a different tenure, with different skills, and with different interests. The number of important decisions taken by government is so great that ministers cannot possibly hope to give their attention to, or even to understand, all of them. There is also a division of power and of interest between ministers and the members of Parliament who compose the legislature. If the functions of the House of Commons were truly formal, if it were a rubber-stamp, then it could hardly be said to share in rule-making. But this is surely not yet the case. The existence of an opposition party in the House of Commons, and its role of making a constant appeal, with the next election in mind, to the public, imposes upon the government the necessity of defending its measures in Parliament not as a mere formality but as a genuine attempt to convince; not to convince the Opposition—that could hardly be hoped for—but to convince interested groups, the more perceptive elements in the electorate, and, most important of all, its own supporters in the House. Thus, although governments do not expect, and very rarely meet, defeat in the legislature, they do not do so largely because the political problems have been ironed out as far as is possible in the consultations, before the introduction of legislation, in the party, with interested groups, in the departments, in the government itself, and in a few cases in the cabinet, or between the Prime Minister and a few close colleagues. Yet this whole complex process of consultation is dependent upon the structure and constitutional powers of the House of Commons. It is the ultimate reality of that body which imposes this whole process upon the government. Let us not decry that institution because it no longer makes laws, for it imposes great restraints upon the way in which they are drafted. Not the least important part of its structure is the rule which excludes holders of offices-of-profit (principally civil servants) from membership of the House. To take this point to extremes, it would be a very different system of government if all members of the House were ministers or civil servants; then truly the function of the House in regard to rule-making would be purely formal. It is this consideration that makes the further increase in the number of ministers in the House of Commons inimical to our constitutional traditions and interests. There is thus a real sense in which the ministers, the civil servants, and the House of Commons share the rule-making power; furthermore, this is a body of men which is, by law, differentiated into three distinct but overlapping groups, and it is legal rules which, by helping to minimize the importance of party among the civil servants, and by helping to ensure a two-party system in the Commons, place limits upon the ultimate power of a single political party or its leaders in the exercise of the rule-making function. It is in this sense that a “partial separation of powers” is still the central principle of the British system of government today. This should not lead to complacency, for, as will be argued later, it is the balance between these elements which should concern us.

Thus there are rules, in this case formal legal rules, which ensure some basis for a functional view of the rule-making power in Britain. These formal rules, however, need support; for, as we have seen, it is very difficult to ensure the primacy of the formal rule-making function when all elements of the government make rules in one sense or another. In Britain this support has been provided by a second set of informal, extra-legal rules of behaviour. Among the bureaucrats, the judges, and also among ministers, there has been a conscious attempt to maintain a distinction between what they have been taught to regard as their own primary functions and the primary functions of other officials, although they might agree, if pressed, that they exercised subordinate rule-making functions as an essential ingredient of their primary function. Thus judges would argue that their main function is the interpretation of statutes, and that although this involves rule-making they will normally attempt to subordinate this rule-making function to that of Parliament. Civil servants could apply policies in such a way that the reverse effect is produced from that which was intended by Parliament, but it is normally assumed that they will subordinate their activities to the statutory intent. Thus professional loyalty, or integrity, the acknowledgment that a certain “function” is their primary concern, is an essential ingredient in the attitudes of ministers, judges, and administrators in the constitutional State. Certain rules of behaviour, internal to the group concerned, are as essential as the external rules imposed by law. The “internal” rules are not, of course, internal in the sense that they are secret, or that they are known only to the group concerned. On the contrary, they are more likely to persist if they are publicly upheld as a “code of behaviour” which distinguishes the group from the common herd. The exemplification of this attitude is to be found in the office of Lord Chancellor in Britain. This office is often quoted to prove that there is no separation of powers in Britain, because this official performs functions in the legislative, executive, and judicial branches of government. This is not, however, a very strong argument, for the Lord Chancellor holds a unique office. It does nevertheless illustrate the importance of internal rules, for this man, when acting as a judge, is expected to show impartiality, and that expectation is enforced by the attitudes of the members of the legal profession, who would quickly denounce any attempt to use the office for purely party ends. Again, there is no fusion of power here, for the “internal” restraints upon the Lord Chancellor are dependent upon his position in a profession the vast majority of the members of which operate outside the government machine.

In the United States the more rigid application of rules to attempt, however imperfectly, to maintain a distinction between those who make, those who apply, and those who interpret the rules, has produced considerable difficulty in the working of that system of government, and many Americans are today impatient with the restraints it imposes. In Britain the partial separation of the personnel of government has been reinforced and buttressed by the sets of internal rules which govern the behaviour of politicians and officials. This comparison is interesting. Americans who find the externalized rules over-restrictive tend to look to Britain as an example of what can be achieved without such rigid rules. But do they, even the behaviourists among them, realize how vital these internal codes of behaviour are, and how far they have in the past depended upon the élitist character of British public life? Without the formal restraints of the Constitution would America have developed, or indeed has she yet done so, a set of internal checks? America has never been a society which was truly run by an élite except perhaps during the years immediately following the adoption of the Constitution in 1789. It was the democratic nature of American society, when compared with that of Europe, that concerned Madison and his friends in the Convention. There was a difference, in their view, between an élite with certain standards and a democratically maintained oligarchy. On the other hand, those who in Britain deny the importance of constitutional rules, or internalized codes of conduct, forget how significant these have been in a country which has always been governed by an élite. If Britain is moving into a more democratic age, one in which the old élite which dominated civil service, judiciary, and ministerial positions alike, is coming to the end of its period of dominance, it may well be that the assumption that rules can be further slackened will prove a danger. The main point of this comparison is that formal rules and internal codes of behaviour may be in part alternatives and in part necessary counterparts.

The usefulness of functional analysis in these terms, it is suggested, is that only by this means can the reality, in some form, of a government by law be maintained. Once the external and internal restraints of the idea of functions and the rules they imply have gone, what else remains? Nevertheless, this does not imply the outright acceptance of the functional categories of Montesquieu or of Almond. The idea of a rule-making function developed above is considerably different from the older idea of the legislative power. But dissatisfaction with the old triad of “powers” goes further than this. In the first place we suggest that there are two levels of the functions of the political system, the primary and the secondary;3 furthermore that at the primary level there are four, not three, basic governmental functions. There is a primary level of functions, which is concerned with the relations between the government and citizens or other governments. Then there is a second level which is a matter of the internal relationships of the parts of government. These two levels are by no means discrete, and it would be difficult to delineate their boundaries in practice. Yet for analytical purposes it is clear that there is, for example, a considerable difference between a legislature concerning itself with a law for licensing dogs and the same legislature engaged in creating an independent judicial system. The forms of legislation may be the same, but the realities are very different.

‡At the primary level there would seem to be four major functions embodied in the working of Western systems over the centuries—rule-making, a discretionary function, rule-application, and rule-interpretation. The importance of rules, and, therefore, of the rule-making function, to a constitutional State has been sufficiently stressed above, but it must be made clear that in the most constitutional of States there must be a discretionary function which is largely free of pre-determined rules. The threefold formulation of the functions of government left no room for the prerogative or discretionary powers, and in the historical application of the extreme theory of the separation of powers the “rule of law” was so strongly asserted that no discretion could be left to governments. But this was an extremely unrealistic view of government both in the eighteenth century and the twentieth. Today in democratic Britain the importance of the prerogative powers of the Crown, especially in the fields of foreign affairs and defence, control of aliens, and internal order, is still recognized. That these powers, exercised by ministers, often provoke the most heated controversy, is a reflection of the fact that such powers are not, and cannot easily be, subjected to prior rules. Of the Constitution of Virginia of 1776 Jefferson said it was not conceived that any power could be exercised that was not subject to law, and the President of the United States was given no “prerogatives” in the Federal Constitution. Yet Jefferson, himself so concerned with the purity of the functional division of governmental power, when in office soon acknowledged the importance of a reserve of discretionary power. The President of the United States has, over the years, developed powers, particularly in the field of foreign affairs, which make him far more than an official who proposes and executes laws, and the courts have been either unable or unwilling to check the evolution of this powerful discretionary authority. The acknowledgement that such a function exists gives added weight to the discussion of control in constitutional systems.

Almond uses the term rule-adjudication, instead of the older judicial power, and this is a most valuable step away from the confusions implicit in the older term, but he does not expand on the nature of this function, and implies that, except for its structural connotation it is co-extensive with the earlier usage. However, the confusions which have surrounded this concept in the past need to be explored and if possible cleared away. The history of this question has consisted of a constant dialogue about whether the judicial power is a distinct and separate power or whether it forms a part of the executive power. On the one hand it has been argued that the judiciary applies the law equally with civil servants, but through a different procedure. On the other hand it has been argued that the distinctive function of the judges is that they decide disputes, whereas civil servants simply “administer.” Thus Montesquieu wrote of “the power of judging.” The longevity of this dispute is explained by the fact that, although both sides of the argument seem to have some validity, the language used in the discussion has never been clear enough to reconcile the two points of view.

There are two elements here—the functional element and the structural. If we study the work of the courts we can see that in the general sense they are applying the rules made by the legislature or by other courts. At the same time, by the creation of precedents they also make rules, which in normal circumstances they consider to be subordinate to, or consequent upon, other broader rules. They do, of course, settle disputes, but this really is a task of government, as is that of punishing criminals, an end product, such as making roads or fighting wars. Thus we can say that the courts apply the law, but they do so in a special way, through a special procedure. The institution of a jury which determines matters of fact is a specific procedural way of safeguarding certain values in the application of the law, for these judgements of fact could be determined by a clerk in an administrative office; but the matters under consideration are thought to be so important to the individuals concerned that there must be a special procedure to determine them. The idea of a “dispute,” which is often said to characterize the exercise of judicial power, is hardly very important. Many of the “disputes” which come before the courts are not really disputes at all, but are arranged to look like disputes in order that the judicial procedure may be applied to them. On the other hand, many matters which are decided in administrative offices are just as much disputes between parties as are the matters decided in court, or at least could be formulated in this way. There has thus been a continuous sifting-out over the course of history of matters which it is considered should be decided by one procedure rather than another. The gradual evolution of the King’s “courts” can be seen as the movement from the position where all business was dealt with in a judicial fashion to one involving a “division of labour” and “specialization,” but not a specialization concerned merely with “efficiency”—rather one concerned with placing emphasis upon different values.

Thus far, then, it seems that the courts do indeed merely perform the rule-application function, but in a different way from administrators. However, this is to ignore an essential aspect of the history of judicial machinery in Western constitutionalism. The courts also perform the function of stating authoritatively what the law is. Whenever the meaning of a rule is called in question the judge must make a binding interpretation of it. Interpretation is an essential step in the application of any rule. It is performed also by policemen, prosecutors, and civil servants at every stage of their work. Each of them interprets the law and then applies it as he understands it, although these two stages will not always be consciously distinguished from each other. The difference between these interpretations and those of the judge, however, is the authoritative quality of the judicial interpretation, whereas those of other officials, although usually accepted as valid, are in principle subject to review. The importance of this distinction cannot be lost sight of in the constitutional system of government, for unless we are prepared to allow the administrator to have the last word on the meaning of the law, some independent check must be maintained upon the interpretations he places upon it. If we were to lose sight of this fact we should indeed live in a society in which the bureaucrat, however benevolent, had the last word.

The judicial procedure, therefore, encompasses the application of the law in two major ways which are different from that of the Civil Service—the facts are ascertained by a special procedure, the law is announced in an authoritative way, and, of course, a single judge may be entrusted with both these functions when a jury is not considered necessary. The reason for the independence of the judiciary, therefore, and incidentally of juries, is not that they perform a judicial function, an expression to which it is very difficult to give a precise meaning. The argument for the independence of the judge is that in performing his function of rule-interpretation he should not be subject to pressure that would cause him to vary the meaning of the rules to suit the views of the persons affected by them, and that in ascertaining “facts” he will not be influenced by considerations of expediency. It is an essential element in the maintenance of that stability and predictability of the rules which is the core of constitutionalism.

We may therefore sum up the primary functions of government as rule-making, a discretionary function, rule-application, and authoritative rule-interpretation. These functions are not closely tied to particular structures in the constitutional State, but the history of constitutional development is the history of the attempt, often hesitant and vague, to articulate government in such a way that a particular structure plays a dominant or important, but not exclusive, role in the performance of a given function. There has been, therefore, a conscious and deliberate attempt to articulate structure and function in a way which would reflect certain values in the operation of government. Although it is impossible to develop a thoroughgoing separation of functions of the kind that the pure doctrine of the separation of powers demanded (and if it were possible it would be undesirable), this does not mean that there is no importance in the attempt to assign the primary or dominant concern with the performance of a particular function to one agency of government rather than another. The whole history of the doctrine of the separation of powers and its related constitutional theories is indicative of the fact that neither a complete separation nor a complete fusion of the functions of government, nor of the procedures which are used to implement these functions, is acceptable to men who wish to see an effective yet controlled use of the power of governments.

At the secondary level the functions of the political system in the constitutional State are even more different from the traditional functional categories. At this level attention has focused upon two functional requirements, control and co-ordination. Men have concerned themselves with these concepts, particularly the former, since the beginning of recorded history. In earlier centuries great emphasis was placed upon the control function, and the whole panoply of mixed government, the separation of powers, the balanced constitution, and checks and balances, was devised in order to ensure the discharge of this function. However, there were differing emphases in the approach made to the problem. The theory of the balanced constitution and its derivatives embodied the concept of the internal checks to power obtained by balancing the parts of the government against each other, whilst the more democratic expression of the pure separation of powers looked for the external control of the various parts of the government by the people. It was considered necessary to divide government to weaken it, and to ensure fairness in its operation, but to subject all its parts to direct popular control. There were many combinations of these philosophies of balance and of popular control, and the vocabulary of one was sometimes transferred over to the other, but they represented the two logical extremes of the approaches to this problem.

Yet why should there be a problem of control in the mass democracies of the mid twentieth century? Does not the establishment of universal suffrage, and of free elections, remove the need for these elaborate ideas about the control of government? It would certainly seem that the electoral system should be a prime means of control in such a State, and clearly it is of the first importance, but is it an adequate, a sufficient means of control? There are, of course, certain technical deficiencies in particular electoral systems, such as the fact that the party gaining the most votes in a British election may quite easily lose that election, but presumably these are matters which could quite easily be put right if the machinery were properly adjusted. Such problems aside, however, it is rare for even the most democratic of thinkers to place his faith in a thoroughly democratic electoral system. Thus in Britain the argument, of all parties, against a thoroughgoing system of proportional representation is that “democracy” is only one of the values, although an extremely important one, which must be taken into consideration. There is also a need for strong, effective, and stable government, and this would be endangered if “the people” were to be represented in their infinite variety. Furthermore governments must have a degree of independence of popular control, they must be given the chance to exercise leadership rather than, merely passively, to follow public opinion. Thus elections must not be too frequent, the electoral system must be tailored to fit the needs of governments as well as of electors, and governments cannot be fully representative, for this would destroy their unity and effectiveness. It becomes obvious, therefore, that the electoral system is not, and cannot be, the sole means of control in a democratic system. It is a spasmodic, and a rather crude, mechanism for the control of government, although obviously its over-all psychological impact upon politicians and officials is enormous.

Are these deficiencies in the electoral system counterbalanced by the structure of political parties and pressure-groups? These organizations are also representative, they exercise an influence upon the government and the decisions it takes. They are the link between the people, who clearly cannot govern directly, and the government, and therefore an essential channel of control. They serve to select leaders and formulate choices in a way which the electoral machinery itself cannot be expected to do. They are a means by which popular views about the aims of government are gathered and registered. The organizational apparatus of the party structure is an essential part of the political system, for it is difficult to see how the necessary choices that must be made could be formulated in an intelligible way without the channels of communication between electorate and government which the parties provide. Political parties, therefore, in a mass society, are an essential part of the machinery of government and a prerequisite of control in a democratic society. The argument for not having a fully “efficient” electoral system is the obverse of the argument for having an effective party system.

Yet, like the electoral system, the party system is by no means a perfect instrument for the control of government. The electoral system, for good reasons and for bad, can and does distort the expression of electoral opinion, for example by preventing the proliferation of parties, or by gerrymandering. But by comparison with the party system the electoral system is an extremely neutral instrument of popular control. Political parties are not merely channels through which opinions are expressed and co-ordinated into simplified choices between men and ideas. They are also organizations through which individuals and groups hope to change opinion, to use popular support for their own ends, to create situations, or to wield power rather than to control its exercise. Political parties, by their very nature, not only link electorate and government, they get between the electorate and the government, and they use the authority of government to attain sectional aims. The studies of leadership and oligarchy in political parties demonstrate that political parties are not neutral instruments of control, and that at the extremes they can be most effective instruments for the abuse of power rather than its control. They must, therefore, themselves be subjected to control, as indeed has been the case in this century with the various laws relating to party finances, and to primary elections.

The argument for interest or pressure-groups is basically the same as that for parties. Through their representative and informative roles they perform an essential control function, but the sectional nature of their membership and aims makes them very suspect as an instrument of control. Nor has it been demonstrated that, although a particular party or group may be only a partial and biased channel for the expression of opinion, the sum of party and group interactions produces an equilibrium in which all points of view and all interests are given adequate weights. This hypothesis, which is the basis of the behaviourist philosophy of democratic government, of which their rejection of the significance of political institutions is taken to be a corollary, is based upon a number of implicit assumptions which cannot be examined at this point; suffice it to say that it does not seem to the present writer that they make out a case, or that it can be shown that the resulting equilibrium is not in large part dependent upon institutional controls.

There are other important structures in the modern State which perform control functions. The press, in the widest sense, radio, and television, are, like parties and groups, essential to the dissemination of information and the collation and expression of opinion. The influence they exert over the formulation of government decisions is undoubtedly a matter of great and growing importance. Yet again, however, we have the two-edged nature of this instrument of control. It is not, and cannot be, a wholly representative or neutral channel. If publicly-owned, the media of communication may become the organs of particular parties or groups who dominate the government; if privately-owned, these media will never be merely the channels for the expression of opinion, but will be used by groups or individuals to shape that opinion. Here too there will, in a free society, be in operation a system of countervailing power, in which the existence of one strongly representative section of communication calls into existence an opposing section. But again, the inequalities of wealth, and organizational power, will ensure that the representative character of these media is by no means perfect.

Thus the argument that the advent of universal suffrage has removed the necessity for control of governmental agencies is certainly not acceptable. There have grown up new and powerful means of controlling government, but like the earlier mechanisms of control they are not neutral instruments, but organizations which must themselves be subject to control. Indeed, there can never be a “neutral” control system, for we must never lose sight of the fact that these “controls” are not pieces of machinery in the mechanical sense. The mechanical analogy is a dangerous one. They are all, without exception, patterns of behaviour, they are all procedures operated by human beings, and they can never be neutral. This, in the last analysis, is the justification of the idea of balance, of setting organizations, government agencies, and groups against each other to provide a means of preventing the control mechanism from taking over and becoming the controller. Control of government can never be a one-way channel, for this will always mean that one group of persons will gain control over others. We can echo Madison here: “If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary.”4 This fact has been increasingly recognized in this century, with movements being instituted to establish a degree of governmental control over political parties, pressure-groups, and the media of communication.

The function of control, therefore, requires in some fashion a notion of balance, whether this balance is expressed within the government machinery, between government and people, or between the media of control themselves. The historic notion of balance cannot be jettisoned at the present stage of development of human institutions. We must reformulate the earlier theories so that they no longer seek a balance merely between executive, legislature, and judiciary, but so that they encompass also those essential parts of modern government, political parties, pressure groups, press, radio, and television.

Thus the historic problem of the control of government remains, in spite of the transformation of the forms of government from monarchical or aristocratic or mixed systems to the modern systems based upon universal suffrage. Yet it would clearly be impossible to leave our analysis of government functions at this point unless we ignored the whole development of political theory since the early nineteenth century. At the secondary level of the political system there is another function, that of co-ordination. It might be thought that this function could be subsumed under control, as it is in fact the function of ensuring that the government works in an effective, coherent way towards the achievement of the goals set it by society. By integrating this conception into that of control we might rid the latter concept of its negative qualities, and forge a new concept of control with positive aims in view. It would be very difficult, probably impossible, to prove either that there was, or that there ought to be, only one secondary function in the modern Western political system, that of “positive control,” rather than the two functions of control and co-ordination. In the last analysis the choice between these alternative formulations is a value judgement—one either places great emphasis upon, and perhaps fears, the tendency towards the abuse of power in political man, and seeks to check it, or one believes that the divisive elements in society can gradually be eliminated, either by “totalitarian” methods or by arriving at a version of the consensus society in which brotherly love rules all. Without being unduly pessimistic the former view of human nature is the one which seems to the present writer to be nearer the truth, and therefore the twin concepts of control and co-ordination, with their potentially conflicting aims, seem more useful than a rather forced view of the unity of purpose in society.

Furthermore, the function of co-ordination might be described as the twentieth-century function of the political system par excellence. The concern of government with new aims has made this the essential function for attaining the ends of modern society, and has often reduced the primary functional mechanisms to mere tools for the attainment of ends other than those they were originally intended to serve. This function has largely been performed by political parties, and by those specialized mechanisms which have grown up in this century—the Executive Office of the President, the modern cabinet system in Britain, with its committees and secretariat. The great importance of this function and its relation to the nature of modern government will be taken up again at a later stage.

The very success of this adaptation of the structure of government to emphasize the importance of a function which is supremely important in the technological, social, and international context of modern government helps to point-up the problem of control as distinct from co-ordination. For the view, associated with writers of the Progressive era, that in the last analysis it is the “people” who must exercise the function of ensuring unity among the parts of government, is wholly unrealistic. The “people,” through the electoral machinery, may set the general pattern of aims for government, but the combination of aims which they choose will have been “assembled” for them by a political party, and even this party programme will normally have relatively little coherence, nor will it give a clear picture of the way in which the tasks of the government will fit, or be fitted, together. In Britain, where party programmes are relatively coherent and are intended to offer the voters a co-ordinated set of programmes, governments, whether Labour or Conservative, have made clear that the actual pattern of government action is a matter for their decision and not that of the party or its members as such.

The differing balance that is struck between the functions of control and co-ordination constitutes the essential difference between the American and British systems of government today. At this point of time, in spite of the pressing problems of the modern world, it seems inconceivable that Americans would ever tolerate that degree of continuously co-ordinated action that the British system today makes possible, or that they would relinquish the control processes that have been given up in Britain. This is not merely a matter of history, it is in large part the reflection of the fact that Americans still do not have an instinctive trust of other Americans. America is still only self-consciously a nation; one part of the country is still unsure of what another part will do, and still has interests which distinguish it sharply from the others. Undoubtedly, as American self-confidence as a nation increases, there will be a greater demand for co-ordination, and consequently a lesser emphasis upon control. In Britain the danger is that the emphasis upon co-ordination will be taken to the point where effective control disappears altogether.

Thus again we return to the idea of balance, a balance between the functions of control and co-ordination. It is important that particular structures should combine the performance of both functions, and that no single structure should be solely responsible for one or the other of them. It is impossible to state any absolute values for either of these, or to set unchanging boundaries between them, only to say that they must be “in balance” and that neither should ever eclipse the other, unless and until that view of the brotherhood of man of which we wrote earlier becomes a reality.

This functional analysis has represented both an attempt to describe and an attempt to draw out the implications of Western constitutional thought and institutions. It illustrates the complex interaction of purposive intentions in the history of constitutionalism with the cold hard facts of organizational needs. Thus the desire to establish the primacy of a rule-making function comes up against, but is not wholly defeated by, the necessity of giving government a wide discretion in certain areas. Furthermore, the functional analysis we have presented here, important though it may be for understanding the over-all character of Western systems of government, does not, prima facie, have very much to say about actual institutions or structures. There are intimations, no more. Clearly a monolithic structure is ruled out, for the idea of balance, and the very concepts of control and co-ordination, in the sense used here, would be meaningless. But if the analysis suggests that there should be a number of structures, it does not say how many, and it certainly does not suggest that each structure should perform only a single function, for the one thread which runs throughout it is the insistence upon the necessity of the performance of more than one function by each structure. Let us turn, therefore, from the analysis of function to look at the organizational structure of modern Western democracies, and their evolution, for further clues to the solution of our central problem.

In general it would be correct to characterize the history of Western institutions to the end of the nineteenth century as a gradual evolution into three great branches or departments of government. It would be impossible here to justify this statement in detail, or to make all the necessary qualifications of it. Furthermore, in the present century this evolution seems to have ceased, or, rather, institutional development has taken on a new and more complex pattern. It is sometimes argued, by Dahl for example, that this development was really a matter of division of labour and had nothing to do with the separation of powers, and certainly the division of labour has had a great deal to do with it. It has been found that certain things can be done more efficiently in certain ways—but what does “efficiently” mean in this context? We have no single criterion, such as output, or profitability, that the economist can apply to this concept. Division of labour has in fact very complex roots when applied to the development of political institutions. It always begs the question “division for what?” and although the answer involves certain technical questions of how things may be done most expeditiously, or more cheaply, it involves a great deal more than this, in particular the recognition of the importance of certain values in the development of Western political institutions.

Let us begin by looking at the building-blocks of organization, in order to suggest some of the factors involved in this institutional development and its relation to function. Any organizational structure is composed of two elements—hierarchy and collegium. At the extreme there is the concept of a perfect hierarchy, best typified in practice, perhaps, by the military organization of a regiment. At the other extreme there is the collegiate body, in which, apart from a presiding officer, the organization is wholly horizontal in character, and perfectly democratic. Perhaps the governing body of an Oxford College is a good example of this type of organization, and so is the prototype Western parliamentary assembly. In practice, of course, these extremes are rarely realized; most organizations are a combination of hierarchical and collegiate elements, with so many possible variants and combinations that the potentialities for experiment are almost limitless. Nevertheless, these two basic structures exist as the poles of organizational structure, embodying very differing characteristics, which in turn lend themselves to the furthering of very different value-patterns. The hierarchical structure has as its major characteristics an authoritative chain of command, unity, and expedition, whereas the collegiate structure involves lengthy debate, divided views which may prevent the taking of decisions, and leads almost inevitably to compromise solutions. These characteristics are, of course, the characteristics one might attribute to the “ideal types” of hierarchy and collegium; undoubtedly collegiate authorities do on occasion act speedily and effectively, whereas hierarchies can in practice become incapable of producing an effective decision; but these are the attributes which the two types of organization in their pure state may be expected to display. Furthermore, the two structures, again in their pristine types, embody very different possibilities of representation. At the one extreme the hierarchical structure is wholly devoid of representative content, for the head of the hierarchy, organizationally at least, is a complete despot, whilst at the other extreme all views can be represented, all arguments aired, all interests can be given their due weight. These characteristics of the two poles of organizational structure come to represent differing value-patterns, and in organizational terms to represent the aspiration towards different ways of taking decisions. The most spectacular confrontations of these ideas, divine right versus parliamentary supremacy, totalitarianism versus representative democracy, involve issues of the greatest ideological and historical significance; but it is also true that at a less spectacular, but nevertheless very important level, the day-to-day operations of Western systems of government can be seen in terms of the tensions between these two organizational types, the continual choice between the values of the one and the values of the other, the attempt to combine the speed and efficiency of the hierarchy with the information and consent which are to be obtained from the collegium.

There is thus a built-in tension between these two types of organizational structure which, however they are modified, can never be wholly removed. This “tension” is the organizational basis of control in government, and it forms the natural and obvious peg on which to attempt to hang functional and procedural distinctions.

The progressive evolution of the great branches of government can be seen, therefore, both as the evolving conflict between differing value-patterns in the way in which decisions are taken, and as a result of the evolving conceptions of functions of government. That is to say that, even if these functional classifications, with their connection with the idea of the rule of law, had not developed, a conflict between the making of effective decisions, by one man or by many men, is built in to the nature of human organizational structure. In fact, these two factors seem to a considerable extent to coincide—that is to say the collegiate organization and the making of general rules seem a priori to be closely correlated, and the hierarchical organization and the application of law seem to be well suited to each other. Certainly most Western theorists have agreed, no matter how sharply they have disagreed on other subjects, that all decisions should not be made by a single man whose word is law, and that all the tasks of government should not be performed by a representative assembly. Except in revolutionary periods these extremes have been excluded from the range of possibilities, so that there must, in some sense, be a functional division at the root of government organization. It is this connection, however difficult it may be to make it precise, between organizational structure, functions, and values, that gives to the separation of powers or rather to its central theme, that indestructible quality we found throughout its history. It is a crucial relationship, to which we continually return, no matter how complex the political and governmental machinery becomes, no matter how intricate the tasks of government. It is here that we reach the abiding core of truth in the idea of the separation of powers, and we can understand also why, throughout the history of Western thought, from Marsilius to the present day, there has been the continual tendency for writers to insist that there are only two functions of government, functions which seem to correspond so neatly with the “natural” tendencies of organizational structure. In fact each of these structures came to perform not one overriding task, but a number of them. Thus the “legislature” is associated in separation of powers theory with the “legislative power,” but its representative nature ensured that it would in fact be associated with all those tasks, such as control of finance, administrative oversight, redress of grievances, and deliberation upon matters of general significance even if legislation was not involved, which seemed relevant to a body with this particular representative structure rather than another. The older term “parliament” better represents the nature of this body than the more modern “legislature.”

The structure of judicial organization, however, has never seemed to fit very neatly into the simple functional classification of those who wished to see all actions of government in terms of a simple psychological theory of willing and acting, for this left no room for the complexities of judicial organization. The judicial system is in fact the clearest indication of the impact of purposive procedures upon organizational structure. It is the expression of the determination to ensure that certain values are given priority at the expense of expediency or speed in the performance of certain types of governmental tasks. It represents the conscious effort to combine the values of different types of organization in order to achieve particular aims. Historically, English institutions have never approximated very closely to either of the extreme poles of organizational structure, although Tudor government on the one hand, and the Long Parliament on the other, have perhaps come fairly close. They have always been a combination of the two types of organization—the King in Council, the King in Parliament—and the tension between those hierarchical and collegiate principles has been a great theme of British constitutional history. Although this clash of the two basic organizational types has been a dominant note of Western constitutional development, perhaps the most interesting, and the most successful, developments have been the attempts to create new structures which combined the advantages (and of course the disadvantages also) of both the extremes. The evolution of the judicial system, with its combination of the King’s judge and a people’s jury, was considered by many English and American writers up to the end of the eighteenth century as perhaps the most important institutional development that safeguarded the liberty of the individual. The parliamentary system, in its mid-nineteenth-century heyday, was seen as the great achievement of a harmonious relationship between the two potentially opposed principles formerly embodied in the ideal types of absolute monarchy and Long Parliament. For the middle-class proponents of harmony in the nineteenth century this system, with its delicate balance, and its internal compensating mechanisms, was the final answer to the centuries of strife between two apparently incompatible forms of organization.

The late nineteenth century and the present century have seen, however, the destruction of this dream of perfect balance and perfect harmony, for harmony suggested an equal, not a subordinate, relationship between the parts of the system. The mid twentieth century has attempted to evolve its own answer to this ever present tension, replacing that harmony of the system of parliamentary government with new forms. To understand this we must look briefly at the main trends of government organization in the past eighty years.

In the first place the twentieth century has seen the re-emergence of the hierarchical principle to a dominant position in government organization. In Britain, first the cabinet rose from a position of rough equality with the House of Commons till it led and dominated it, and then the growth of the power of the Prime Minister lifted him up far above his cabinet colleagues in power and prestige. At the extreme this has been labelled “government by Prime Minister,” and it has been argued that the cabinet as an institution has joined the monarchy and the House of Lords as a “dignified” rather than an “efficient” part of the Constitution. The English system is now portrayed as the Prime Minister, supported by a few close associates, directing his ministers, controlling Parliament, and through the whole hierarchy of cabinet committees and sub-committees maintaining a grasp upon the administrative machine: an elective monarchy, in fact. This, no doubt, is an exaggeration. The limits upon the Prime Minister’s power clearly vary very much according to circumstances, and the usual exemplifications of his great power are drawn from the field of foreign affairs and defence rather than from domestic politics, where he has much less freedom of manoeuvre. Still, the general picture is nevertheless one of the growing dominance of the hierarchical principle. This is the result of the greater demand for emphasis upon co-ordination noted earlier, and for the speed, despatch, and relatively purposeful activity of a single man rather than a committee or assembly. The delicate mechanisms of the nineteenth-century theory of parliamentary government have been transformed into very different instruments. Ministerial responsibility is now little more than a formal principle used by ministers to deter parliamentary interference in their affairs, and the power of dissolution has become simply a tactical weapon in the hands of the Prime Minister to enable him to choose as favourable a date as possible to fight an election. The enormous growth of presidential power in this century in the United States, and the Constitution of the Fifth Republic in France, reflect the same tendencies, with the same basic forces at work, although the different political structures of those countries have modified the methods adopted and the detail of their application.

At the same time that this emergence of the hierarchical organization as the major force for initiation and co-ordination was taking place, another, contradictory, development was in process. There was a fragmentation of the structure of government in an attempt to combat the growing importance of the hierarchical principle, an attempt to modify its force. This was most marked in the United States, where the development of the independent regulatory commissions took a large area of administration out of the direct control of the President, and thus provided some form of compensation against his growing power. The wide use of these independent or semi-independent agencies in the United States illustrates the emphasis placed upon the control function in that country. It was an almost automatic response to the relative decline of Congressional power. Although in Great Britain we find some tendencies towards this type of fragmentation—in the semi-independent status of the National Assistance Board for example—the emphasis upon the co-ordination function has been so strong that the fragmentation of structure we find in the United States has not occurred to the same extent.

The growing dominance of the hierarchical structure has met with a rather different reaction in Britain. Instead of attempting to frustrate the hierarchy by imposing external checks upon it, there has been an increasing tendency to build in internal restraints upon its action, and to create in fact a new combination of the hierarchical and collegiate forms which might achieve some of the same ends that were implicit in parliamentary government, without the dispersion of effective decision-making power that the initiative of rule-making power in the legislature implied. In the early part of the century the relative decline in the importance of Parliament led to proposals which suggested the creation of some form of corporative parliament in which interests would be represented rather than geographic areas. This formal proposal never gained acceptance in Britain, but much of its spirit has been realized in other ways. In the first place, the hierarchy of ministers and civil servants was encrusted with an enormous outgrowth of advisory and consultative committees, which provided information and expressed opinions of interest groups. The practice of formally consulting interested parties before important decisions are taken has virtually become one of the new “conventions” of the Constitution. In 1962 this attempt to democratize the administration advanced an important step further. The problem of obtaining consent for government programmes that planned to maintain restraints upon wages, salaries, and prices led to the creation of the National Economic Development Council, which was a formal attempt to integrate administrative and representative organizations and procedures, so that at least part of the control of the government’s policy that it refused to surrender to Parliament was entrusted to a body representative of only certain sections of the community. This type of development has gone even further in France, where the planning machinery encompasses a much wider consultative apparatus. Indeed, M. Mendès-France has proposed this as a pattern for the evolution of a modern structure of government for France.

Thus there has been an attempt to infuse the administration with representative structures, and to find some sort of balance between expeditious government and representative control within a single structure. At the same time there has also been an attempt to infuse the judicial values of fairness and due process into the administrative procedure. Thus the growth of administrative tribunals, and the use of “quasi-judicial” procedures, represents an attempt to give due weight to the interests of the individual, without destroying the speed and effectiveness of government action. There has, therefore, been a twofold attempt to dilute the hierarchical form rather than to exert an external check upon it, by building collegiate organizations and judicial procedures into the very structure of administration itself. This is a reflection of the continuing desire for the performance of an effective control function, but it also represents a determination to maintain the advantages of the hierarchical form of organization.

It is doubtful, however, if these attempts could be said to be successful. The dissatisfaction with the way in which the control function is being performed has produced the Council on Tribunals to attempt to give some independent supervision of administrative justice, has led to demands for the reform of Parliament to make control of the administration more effective, and to the proposals for a Parliamentary Commissioner to investigate grievances. There has thus been a resurgence of demands for more effective external checks to be applied to the hierarchy of ministers and civil servants, and a suspicion that bodies, which are representative of outside interests, but which work for long periods with the hierarchy, become identified with it in the minds of the ordinary people, and perhaps in their own minds as well. There is the awful fact that a decision which you have helped to make must in part at least be defended by you, and it becomes, in the eyes of other people, your responsibility. This fact has always been recognized by parliamentary oppositions, who have consistently refused to accept any responsibility for, or even to enter into private discussions on, government policy. The need for an external check of some description seems continually to emerge from these situations. It is this fact which, above all else, suggests the need for a Parliament with effective control functions. Parliament has given up any attempt to initiate or even amend rules which are to be made. It must, therefore, control the means by which government is carried on. Only in this way can an external check be applied. We should not allow one-sided appeals to the out-dated vocabulary of the separation of powers to prevent this; to talk of interference in executive functions by Parliament is today the most cynical use of terms that no longer have any real meaning.

The final way of looking at our material is summed up in the term “process.” This is a term which has as many meanings in the literature of the study of politics as has “function,”5 and undoubtedly it is often used simply to give the impression of being modern and up-to-date. It has value in the sense that it stresses a concern with the whole complex of political activity rather than with merely formal elements, for the term “institution” has come to take on the connotation in the study of politics, unlike sociology, of a set of lifeless forms, with perhaps little relevance to what actually goes on in the hurly-burly of everyday politics. The term process can, however, be helpful in ways other than merely demonstrating our up-to-date attitudes. In particular it can point attention to the importance of the dynamic elements in the study of politics. A. F. Bentley used process simply to mean that in politics all is movement, all is flux.6 It is true that the political system is in a constant state of change, nothing ever stands still, patterns never repeat themselves exactly. Yet, as has been argued above, this does not mean that there is no stability in political life, for the political system and the idea of order are inseparable. If all were flux, there would be no possibility of foreseeing the outcomes of political actions, no basis for rational behaviour—in fact, no politics.

Thus, if the concept of process means anything in politics, it does not mean that all is flux. It can help to focus our attention, however, upon the problem of how events move from point A to point B, of how the situation at the end of a period of time is different from, but clearly related to, the situation at the beginning of the period. We have to take into account time sequences when discussing the nature of political institutions; thus institutions, or structures, are patterns of behaviour that persist over time, but they are never exactly the same after the passage of time. What makes them recognizable as institutions is the fact that they show a basic stability and continuity, which allows them to adapt to changing circumstances without losing their identity. At any particular point of time we must concern ourselves with how people are behaving in relation to past patterns of behaviour, and take note of their expectation of how other people will behave in the future. At any point of time the rules explicit or implicit in past patterns of behaviour will be influencing people to continue broadly as they have done in the past, but there will also be an infinite number of variations in behaviour from past patterns, many of them minor and ephemeral, some of them important for future patterns of behaviour. There will thus be a constant potential for change, which will normally be marginal, but which can over a long period alter the general structure very considerably.

If we think of this whole complex of political behaviour as the political process, it becomes in fact coextensive with politics. “Process” really becomes redundant. However, in so far as we divide up the study of the political system into a number of areas, the idea of the legislative process does convey something of the combination of stable, continuing activity which forms the core of the operations of a legislature together with the innumerable ephemeral political acts of those interested in the particular events that are its concern. The stable institutional pattern will operate, must operate, in terms of a set of rules implicit in the behaviour of those who are involved in it, or which may be made explicit in a set of formal constitutional or legal rules, or explained by observers as “conventions” or usages. The more formal, written rules may, of course, become out-of-date, so that they are no longer in practice regulating, or influencing, the behaviour of the actors in the situation. In this case, they will be modified, or superseded, by unwritten codes of behaviour more relevant to what actually happens. It is the unfortunate tendency of revisionists in history or politics, having demonstrated that the old written rules or unwritten conventions no longer have relevance to the practice of a political institution, to assume that such rules never were important, and that no such rules are important today. Yet what has usually happened in fact is that a new set of rules, different from, but usually clearly related to, the older set, is now in operation.

In this wide sense, therefore, there must be a procedure at the heart of every political process. Procedure, which is just another name for these rules, is the distillation of the institutional pattern; not an unchanging, rigid set of rules, but rather a flexible body of precepts which contain the provision for their own amendment. Viewed in this light the political system will reveal many processes and their related procedures—certainly not merely the three, legislative, executive, and judicial, of the earlier writers. There will be a political party process, and an electoral process. There will be a tendency to create new procedures to meet new needs, just as the gradual evolution of three distinct procedures in earlier centuries reflected the changing aspirations of the peoples of England and France, and as the development of such new procedures as those of the independent regulatory commissions, or administrative tribunals, reflected newly emerging problems and the desire for their solution. There is nothing sacred, or divine, about the trinity of legislative, executive, and judicial powers in earlier theory. It is a matter of the procedures which are felt to be necessary to meet current needs.

We have seen, however, that although historically and logically there is no justification for the view that there must be three and only three “powers” of government, there is a remarkable stability in the general articulation of the parts of the constitutional State and of the procedures they adopt. It is here that we may begin to see some of the inter-relationships between function, organization, and procedure.

In the first place the functional characteristics of the constitutional State exercise great influence upon the number of different procedures adopted. The constant pressure to ensure that the rules governing behaviour are explicit and formalized, so that each person is aware, so far as possible, of the consequences of his actions, that is to say the ancient demand that people shall be governed, in Locke’s phrase, by “promulgated establish’d laws, not to be varied in particular cases,” will inevitably tend towards the creation of a hierarchical system of rules, with a single final source of authority and a procedure for testing their legality. The demands for a balance between the co-ordination and control functions will inevitably lead to a limited number of structures and procedures, but a number greater than one. Thus we find continual attempts, throughout the twentieth century in Britain and the United States, to keep the evolving structures of government in a pattern which will ensure that they are co-ordinated and subject to control. The reports of the various commissions of enquiry into government organization in these countries illustrate the pressures which operate to bring about the consolidation of these structures into a single hierarchical structure, subject to the control of representative institutions, with only those exceptions considered necessary because of the values they represent. On the one hand, no means of checking the exercise of power without some form of external restraint has yet been evolved; on the other hand, the pressure in the modern state towards a set of co-ordinated harmonious government policies is such that the integrative tendencies will always be at work. The nature of the basic forms of organization, hierarchy, and collegium will also tend always towards two or more, but not many more, organizations and procedures. There are characteristics of those two forms which make them suitable to certain tasks and to certain modes of proceeding. As we have seen, the most interesting and significant developments have been in terms of the attempt to create new combinations from these basic forms, yet the organizational pressures that tend to polarize institutions remain very great.

Second, it cannot be too strongly stressed that procedures, the rules governing behaviour, reflect certain value-patterns. The way in which things are done makes a very great difference. Men could be condemned to death, and in some countries are, by an administrative procedure. Roads could be built by a collegium determining by vote, after discussion, where every stroke of the pick should be made. The judicial method involving open discussion and an adversary procedure before a jury could be used to determine important questions of foreign policy and diplomacy. The results of allocating these tasks of government to be decided in this way would undoubtedly be disastrous. The present-day procedures in Britain and the United States, and the matters decided by them, have not been evolved by chance; they represent the collective judgement of centuries concerning the way in which certain things should be decided. This is not an argument against all innovation, but it should lead us to enquire into, and to examine the values which these procedures embody, and to look very closely at new procedures, and at the allocation of tasks to them, in order to be sure what we are doing.

Thus procedures, which form the heart and core of every political process, may be seen as the institutional expression of the value-patterns of particular societies. These value-patterns are extremely complex, but broadly speaking the evolution in modern times of three major procedures of government reflected the importance attached to three dominant values in the Western World—efficiency, democracy, and justice. Over the past hundred years, however, a new value emerged which could not be subordinated to these—social justice. It is the concern with social justice which above all else has disrupted the earlier triad of government functions and agencies, and has added a new dimension to modern government. It has resulted in the creation of new structures, and the evolution of new procedures, but its implications go beyond the mere multiplication of values, structures, and procedures; for, with all the difficulties which the functional classification of government that dominated the years 1640–1848 presented in detail, the broad correlation of the three concepts of government function with three structures was a reality. The three values of efficiency, democracy, and justice did, of course, come into sharp conflict, but this conflict could be institutionalized and controlled; it was this conflict that gave the fundamental impulse to the concept of the separation of powers. Thus functional intentions, organizational structure, and the values implicit in procedures combined to give a meaning to this constitutional doctrine. There was a logic behind the apparent illogicality of much of the writing about the separation of powers.

The growing importance of social justice, however, threatened to destroy this logic. This new value cut across the other three in new ways, and it could not result simply in the construction of new structures and new procedures to form a new fourfold separation of powers, although the fragmentation of government structure which took place in the twentieth century reflected in part an attempt to do this. To a considerable extent this was due to the fact that the aims of social justice had to be achieved largely through the same machinery which in earlier ages had been concerned almost exclusively with maintaining order, conducting war, and diplomacy, or dealing with the minimum routine needs of society. But also it was because the achievement of social justice meant more than the distribution of new goods and services; it meant also the control of the economy to ensure full employment, the attempt to secure the incomes of farmers and wage-earners, the control of monopolies, the maintenance of a certain level of public expenditure, the control of the balance of payments, and so forth. The measures needed to achieve these aims cut across the older values; in particular they entrenched upon democratic controls and judicial procedures, and they demanded far more co-ordination of government action than in the past.

Indeed, it could be argued that this new value had not been added to the earlier ones, but had become the value, an overriding factor which did not have to be articulated with the others, but superseded them; to the extent that they could be accommodated to it, they would survive, but no further. The rise of the modern mass-based political party is closely connected with this emphasis upon the value of social justice. In fact the twentieth-century political party is the structure through which this value has been realized, just as in an earlier age the representative assembly was the structure through which democracy was realized. By performing, above all else, the function of co-ordination, by using the primary functions as mere tools, with little concern for the ends they had been fashioned to pursue, and by ensuring the creation of new co-ordinating instruments of government to further their aims, political parties have become governmental structures par excellence. Of course, the other structures of government also perform this function—they are no less multifunctional in this respect than in others—yet it has become the prime function of the political party. Thus the analysis of the correlation of value, function, and structure remains complete, but the picture has changed because this value is seen as superior to the others, because the main purpose of this new structure is to co-ordinate the older structures, and therefore to some extent to subordinate them. The political party process has come to encompass the electoral, legislative, and executive processes, and indeed the judicial process as well. The rise of this new value and the structures which ensure its realization must mean that any facile view of “the separation of powers” is dead.

It is natural that the emphasis upon one value, social justice, and the functions and the structures it entails, should have been so great in a period when the realization of the shortcomings of earlier ages in this respect had become so intense. We have witnessed a revolutionary change in attitudes, and there has been a consequent extreme emphasis upon the new value, and its institutional expression, of the kind which has accompanied earlier revolutions. The overriding importance attached to “democracy” in revolutionary situations at the end of the eighteenth century led to an extreme emphasis upon the power of representative assemblies, only to give way to a compromise between the old and the new when the dangers were seen of erecting a single criterion, a single value, to this dominating position.

In the mid 1960’s it cannot be said that social justice has been accomplished completely in Britain or the United States, any more than it could be said that democracy had been completely achieved at the end of the nineteenth century, or justice completely achieved at the end of the eighteenth. Yet we have perhaps come to the point where we must pause, and turn again, as earlier ages have turned, to the reconciliation of the new values with the old; to question whether one value, however important, can be allowed to exclude others. Human beings are much too complex to be dominated for long by one overriding consideration; they demand a number of satisfactions, usually potentially contradictory ones, if pursued to extremes. A system of government which is to meet these demands, which will respond to a variety of values and their functional and structural requirements, must attempt to reconcile the old structures and procedures with new ones. Control will be important alongside the function of co-ordination, to maintain a balance between differing views of the nature of government; the primary functions will be considered important in order to give expression to older values that we cannot relinquish; even the old concept of the separation of persons in government will be important where it is seen, not as an end in itself, but as a means of maintaining this balance. The task of the twentieth-century political theorist is to place these values in perspective, and to suggest the institutional means by which they can be reconciled.

[1. ]G. A. Almond and J. S. Coleman, The Politics of the Developing Areas, Princeton, 1960, pp. 16–17.

[2. ]In this discussion I am greatly indebted to an unpublished manuscript on administration by Andrew Dunsire of the University of York.

[3. ]This terminology again is taken over from the rather different usage of Professor H. L. A. Hart.

[4. ]Federalist, No. 51.

[5. ]See the discussion by W. Harrison, Political Studies, Oct. 1958, p. 243.

[6. ]See Norman Jacobson, A.P.S.R., Mar. 1964, p. 15.