3rd Revised Draft
31 March 1991
(c) John Barnes
Attempts to encapsulate the distribution of power within the British political system in a single phrase have led to much sterile debate over the last thirty years. Nevertheless it is premature to dismiss the major proponents by reference to the normative nature of the phrases concerned. Mackintosh's principal concern in challenging the use of the term "Cabinet Government" was not normative but analytical [Mackintosh 1962, 1968]. He believed that a change had taken place in the distribution of power within the political system that was being obscured by the use of the traditional term. In effect, he was calling for a new model of the system, nor does it seem chance that he did so shortly after de Jouvenel claimed that the trend in modern state systems was towards the rule of a single dominant figure. [De Jouvenel 1960]
Part of the difficulty in resolving the debate lay in the absence of any systematic examination of the evidence, but a more fundamental problem was the common assumption that the distribution of power was static. The relative ease with which attempts to find a single pattern of power to match all occasions and all times in Government's life could be falsified suggests that the policy process is too dynamic to be captured by a steady state model. Such attempts are best compared with the task of capturing the patterns produced in a kaleidoscope in a single photograph.
The gradual realisation that the policy process was not a single entity and that it might differ in discrete policy areas was liberating, although again constrained by the assumption that there was a "normal" pattern in each area awaiting discovery. It allowed a more realistic picture of Prime Ministers at work, their efforts necessarily and increasingly employed in inter‑governmental relations, the conduct of the economy and the shape of Britain's defences; within the constraints imposed by lack of time or energy, the remainder of their attention was focused on subjects which they thought important or of interest. Inevitably areas were found in which Prime Ministers scarcely figured, except as arbiters in disputes. Such areas varied from one regime to another. Further no attempt was made by the protagonists in the debate to correlate their analysis with that offered by proponents of bureaucratic power [Dunleavy 1990], although it was evident that Government departments had a life and will of their own.
The fact that power can be distributed in a variety of ways within the confines of the policy‑making system is not a cause for despair. There is no reason to suppose that the distribution is random, and clearly it is constrained by the way in which the system is organised, by the operating procedures followed and by what may be called the rules of the game. What political scientists should have been concerned with, therefore, were the circumstances in which one possible pattern was more likely to emerge than any other. What is needed is a precise specification of the circumstances in which and the reasons why a particular constellation of power emerges, constraining policy outcomes. This should be done without normative reference.
As in most social science, law‑like statements are hard to establish, but it is within our compass to analyse the pattern of power and suggest why in particular circumstances one kind of outcome is more likely than any other. It should be possible also to establish those variables that make complete predictability impossible. Further than that political science cannot hope to go.
The six models of the core executive identified by Dunleavy [1990] are best seen as different states of the system. Not only do they fail to encapsulate the whole truth, but they are too static in their characterisation to match up to their heuristic purpose. More precise specification would help, but would lead, almost certainly, to the conclusion that each accurately captures the way in which the system functions at different moments in time and in discrete areas. Given discrete policy areas, different models can be in operation simultaneously. Nor is Dunleavy's list exhaustive. There are cogent suggestions in Headey's rough categorisation of ministries and policy areas [Headey 1974] which can be developed to meet a point Dunleavy makes about the need to specify more completely the nature of the component parts to Whitehall. Process as well as structure needs specification, however. Studies of nuclear decision taking and of energy policy more generally suggest that highly technical decisions give rise to their own particular pattern of decision‑making, while practitioners like Jenkins and Callaghan have spelt out the differences between decision making in the Home Office and the Treasury.
Laski's last published lecture on the subject, given in 1950 [Laski 1951], is as convincing an account of the Prime Ministerial hypothesis as anything produced in later debate. However none of these writers thought that the Cabinet had been relegated to the "dignified elements of the Constitution" and the reason is quite plain: the power of the Prime Minister is a given in the system but it is neither be all nor end all.
Indeed there is a good deal of recent testimony from senior civil servants that the centre of government is too weak. The late Lord Armstrong, a former head of the Civil Service, when talking about British government, laid stress on the fact that it was a "government of departments". In his Reith lectures Sir Douglas Wass, formerly Permanent Secretary of the Treasury and joint Head of the Civil Service, argued that the form and structure and agenda of a modern Cabinet "almost oblige[d] it to function like a group of individuals, and not as a unity", a judgement amply confirmed by Bruce Headey's interviews with Cabinet Ministers from both parties [Headey 1974], and he concluded that such a body not only could be railroaded by a strong departmental minister but that it seldom had the time or opportunity to review and assess the general thrust of the Government's policies [Wass 1983]. In this he echoed Amery [Amery 1935]. To some extent also he was expressing views similar to those put forward by a distinguished Permanent Secretary, Sir Richard Clarke, to his colleagues [Clarke 1971]. Sir Kenneth Berrill, head of the CPRS and later Chief Economic Adviser to the Government, when making a case for a Prime Minister's Department, argued that the sum total of departmental interests did not amount to a strategy and that incoming ministers speedily donned departmental spectacles. Sir John Hunt, Cabinet Secretary from 1973 until 1979, also concluded that Cabinets were not well placed to exercise a clear strategic oversight in relation to the manifold activities of Government, but if they did not do so, there was "an inbuilt risk that decisions may be taken in an arbitrary, uncoordinated and even contradictory manner." The reason ‑ Cabinet ministers were immersed in their departmental work [Hunt 1983]. All in one way or another were confirming Crossman's view that "perhaps the biggest task of a Prime Minister is to stop the fragmentation of the Cabinet into a mere collection of departmental heads" [Crossman 1972].
The case for reform made by these highly placed witnesses, and the precise shape it should take, is not germane to this paper, but their testimony is significant in the emphasis it lays on departmental pluralism with an ill‑equipped Prime Minister as the best person to check it.
In the light of this evidence, there is good reason to suppose that what the Prime Ministerial government model and the departmental government model offer the student are twin poles to the policy making system. Neither affords a true picture of the process, and, like the magnetic poles, they have no real existence. The tension between the two poles, however, is real enough, and is a major feature of the system. Before dealing with its constituent elements, it is probably important to repeat that what has first to be described and subsequently specified is dynamic in nature. Its constituent parts may remain constant over time, but the way in which they interrelate varies a good deal and the distribution of power is fluid rather than static.
The authority of the Prime Minister derives less from his position as the author of the Cabinet's existence and its chairman than from his position as leader of the majority party in the House of Commons. Significantly in a coalition government, the position of the Prime Minister is weakened, particularly if he does not lead the largest party within it, although it should be noted that in such circumstances his right to dissolve Parliament assumes much greater importance than analysts usually accord it.
Where majority governments are concerned, the major variables other than purely personal factors affecting a Prime Minister's authority, and at times even the likelihood of his remaining in the job, derive from his relationship with his party, more particularly the party in the House, and with the electorate. Although the differences between the parties are probably lessening, the nature of the party and its conception of authority are also factors worth noting. The Conservative Party, for example, has a rooted belief in its own right to govern. So long as a few deep‑rooted prejudices are respected, it is relatively pragmatic about its ideology and will respond to a strong lead so long as this brings electoral success. Macmillan was able to secure the acceptance of indicative planning in 1961 and Heath accomplish his U‑turn and the acceptance of a statutory incomes policy in 1972 without many overt signs of revolt. The Conservative conception of a leader's authority, both within the party structure itself and at whatever level the party is in office, follows a common pattern. In essence it is that of a cricket captain and team selector rolled into one. Once elected, the leader is accorded a wide degree of latitude and there is general acceptance that he will determine the strategy, select the team and place the field. This is regarded as the most practical way of running things and the leader is judged on whether s/he produces results. The party defers, and, where a leader delivers, will continue to do so, but it is correspondingly ruthless with those who are seen to fail. Nor does it always wait until that failure has been demonstrated at the polls before acting. There were powerful pressures on Macmillan to go in the aftermath of the Profumo affair in the summer of 1963 and Mrs Thatcher fell victim to a similar panic in the autumn of 1990. Labour leaders have in general proved more secure, and now that the process of replacing them involves both the party conference and a college of electors, it is widely felt that they will remain so.
Although intra‑party democracy and the calls of ideology are more of a constraint on a Labour leader, Prime Ministers from both parties usually need pay little attention to the views of their activists until an election is imminent. However they may have to do so if the Parliamentary party is divided and, more particularly, if the Parliamentary leadership is itself split. For Labour governments the inbuilt majority of the trade unions at Conference is a mixed blessing. For much of its past in government this has been a stabilising factor, but the movement of key unions to the left gave the earlier Wilson governments some uneasy moments, culminating in its outright defeat at the hands of the trade unions and their sponsored MPs in 1969. Tensions were resolved once the party went into opposition by the creation of the social contract, but this in turn meant that the 1974‑79 Government had to accord the trade unions an unusual degree of deference, certainly much more than its Conservative counterpart accords its aligned groups, though not more perhaps than they give certain classes of voter.
Conservative leaders have to go considerable lengths to satisfy their following in the House. In this respect a long tenure of office generates its own strains. The numbers of the disappointed and the sacked mount, and if majorities are large, the disaffected have room to manoeuvre. Eden to some extent fell victim to the removal of the repressive effect of Churchill's immense public reputation on his followers, who at times had been disappointed at his actual conduct of affairs. After five successful years in office, Macmillan's reputation faded with remarkable speed, and Mrs Thatcher suffered from a number of switchback rises and falls in her reputation with the Parliamentary party in the course of more than eleven years in office. Norton has documented fully the ways in which Heath antagonised his followers [Norton 1978]. The practice of annual re‑election, although it took almost a decade for its significance to be realised, has not only made the importance of this relationship more explicit, but arguably has made it much easier for the discontented, often in tacit alliance with the press, to sap a Leader's authority.
Party and the party's conception of leadership are key variables, which interact with, and highlight the importance of two other key variables, the Party's standing in the polls and, perhaps, more important, the Prime Minister's standing with the electorate.
While a Prime Minister's authority and powers of patronage are at their summit in the aftermath of an election victory, a pattern of continued policy success can reinforce Prime Ministerial dominance in those areas in which the Prime Minister chooses to be active. Evident policy failures correspondingly have a potential for damage that can be offset only partially by reshuffles and the process of policy succession. It is no accident that Mrs Thatcher's defeat over the General Motors' proposed purchase of British Leyland's truck division came in the aftermath of the Westland debacle, nor that she found Pym a good deal easier to sack than she found moving Howe from the Foreign Office. If a Government loses its way in by‑elections, local elections and with the opinion polls, the authority of a Prime Minister is substantially damaged and the tendency for colleagues to assert their authority is reinforced by the evident desire of most Prime Ministers in such circumstances to look for increased collegiality out of personal prudence.
Historically the Conservative Party has a tendency to panic when election auspices seem grim. Mrs Thatcher was in deep trouble once the main opposition party established its dominance in the polls and bye‑elections resumed their function of making headline news. The imminence of a General Election or the absence of any agreement on a successor can sometimes ensure a Prime Minister's survival in office even though it does little to restore his authority within Government. As Peter Jenkins noted of Wilson in the aftermath of his climbdown over the proposals embodied in Barbara Castle's White Paper, In Place of Strife, in 1969 the power of the Prime Minister was "sufficient for him to remain in office, but insufficient for him to remain in office and have his way." [Jenkins 1970] Nevertheless Wilson came through a grim period to lead his party into the next election. Attlee similarly survived in 1947. Eden and Macmillan arguably were less fortunate. Both the ERM decision and such straws in the wind as Kenneth Clarke's ability to insist on ruling out education vouchers, it seems as a condition of accepting the DES, were indications that Mrs Thatcher's authority was in a bad state of decay, but few observers anticipated that the Conservative backbenchers would use the process of annual election to destroy her. At least in the case of a Conservative Prime Minister, but more probably Prime Ministers from all parties, continued backbench support proves to be a key variable in their survival.
Good relationships with backbenchers can clearly minimise even if they cannot wholly remove the dangers which arise from a party's electoral unpopularity. Norton [1977] has shown how Heath ran into problems precisely because he had neglected them, and Mrs Thatcher, at least in her earlier years as Prime Minister, appeared to have learnt the lesson. However their acquiescence, if not their support, is also necessary to secure a great many of the policies the Government wishes to pursue. Of course it is possible for a Government on occasion to secure the passage of an unpopular bill: Finer has charted the way in which Heath obtained passage of the Resale Price maintenance bill in 1964, but it is worth noting that an election was imminent, he had the backing of a popular Prime Minister and that, even so, he had to make concessions. [Finer 19xx]. Butt has provided several instances of backbenchers flexing their muscles in the 50s and 60s [Butt 196x], and Wilson's defeat in 1969 over In Place of Strife was in the last resort the product of a backbench revolt. Nor was the Wilson Government able to reform the House of Lords despite frontbench Conservative support. The backbenchers of both sides made its passage impossible. More recently the Conservative Government has lost its Sunday Trading legislation. A good example of a minister forced into a humiliating climbdown on previously announced policy is that of Sir Keith Joseph who in 198x had to abandon his attempt to seek a parental contribution to student fees. Backbenchers can also be drawn into Cabinet arguments. Both in 1947, when the Prime Minister and the chairman of the relevant Cabinet Committee seemed bent on compromise over steel nationalisation, and again in 1967, when the Prime Minister used them to shipwreck the policy of supplying arms to South Africa favoured by some of his colleagues, Labour backbenchers were deliberately involved by the protagonists in Cabinet arguments [Dalton 1962; Brown 1972; Healey 1989]. Of course backbenchers can overplay their hand. The group of Conservative backbenchers in 1980 who asked for more concessions on the immigration rules than they had already obtained from the Home Secretary were deserted by their colleagues who thought them over‑demanding. The most recent example of the process concerns the Cabinet row over child benefit where the arguments deployed were spread by a process of deliberate leaking to the backbenchers and their pressure undoubtedly resulted first in the concessions made by the Thatcher Cabinet and then the change of policy agreed by Major and Lamont. Open dissent, however, is simply the failed face of party management, and every Cabinet minister has a Parliamentary Private Secretary to keep him in touch with feeling in the House on a more individual basis than that afforded by the Whips' network. Cabinet Ministers know even if political scientists do not that they neglect their colleagues in the House at their peril.
At first sight, this might be taken to strengthen the hand of potential leaders within the Cabinet, since each has some following on the backbenches. To a very considerable extent, however, they hold each other in check and are not usually a threat to the Prime Minister's continuation in office. Nevertheless it is not surprising to find that some Prime Ministers, Macmillan for example, have sought to deny key offices to such men. That is not easy to do, and in any case the most surprising figures can develop not only leadership aspirations but some potential as challengers when given the Treasury or the Foreign Office. While Parliamentary standing can never be ignored, and the House of Commons should figure in any analysis as a "stock exchange" dealing in reputations, the main basis of a minister's power in policy making derives from the fact that they head departments of state, powerful ongoing entities, which can organise information and argument to advance their cause.
Much has been made by ministers of both political persuasions of the power of the bureaucracy in terms of shaping the definition of the problems with which they engage, controlling the flow of relevant information, and constraining the options which they consider. Paradoxically, much of this testimony comes from ministers who demonstrably have not been prisoners of the machine. However, the available evidence suggests that while civil servants are adequately socialised on the importance of respecting manifesto commitments and on the need to avoid ministerial embarrassment, they are ready to engross the ministerial role whenever it proves necessary and are tempted to do so even when it is not. Most of them have a deep‑rooted belief in the need for continuity and stability in policy, allied with a sense that their way of doing things is better than anyone else's. While the cruder versions of "bureaucratic imperialism" [Niskanen 1973] need considerable qualification [Goodin 1982], a combination of incrementalism and various bureaucratic imperatives have undoubtedly given the departmental civil servants a major role in the formulation of policy. In a great many areas, their work is simply rubber stamped by ministers. The major problem for even the ablest ministers is shortage of time. An effective minister has to concentrate on perhaps as few as two or three issues at any one time if he is to make his will effective, although it is fair to add that the more thoroughly he is known to go into matters, the more likely it is that his known opinions will be followed in other policy areas.
But before subscribing too readily to the bureaucratic model, students of policy‑making need to consider another side to the relationship. The department needs a strong minister to win its battles in Whitehall as well as to obtain credibility in a wider world. No civil servants are present when ministers battle head to head with the Chief Secretary over the annual increase in the department's expenditure, nor when the Legislative Committee of the Cabinet is considering what bills will be included in the programme for the session. Nor is a minister with standing and/or the Prime Minister's ear a negligible ally in the turf battles which beset Whitehall. The Dame may have been capable of taking on Whitehall when her Housing and Local Government empire was threatened by the newly created Ministry of Land and Natural Resources in 1964, but the game was almost thrown away at the outset before she had time to brief her minister. She had also to involve him to make her victory secure. [Crossman 1973] Since their needs are congruent, a close working partnership often evolves at the top of the department which involves an implicit trade‑off between ministerial preferences and departmental needs. The precise balance of power between a minister and his most senior civil servants remains hard to assess, even by those most closely concerned, and perhaps only the Private Office, if it ever had time to reflect, could judge with certainty. While there are ministers patently carried by their department, such purely ambassadorial figures are rare, and however well‑liked personally, the department is glad to be rid of them. Hard taskmasters like Sandys may be something of a pain in the neck, but in getting their own way, they usually also serve the department's interests well.
Arguably, therefore, the relationship is best seen as symbiotic, and it is the first of two symbiotic relationships, the correct specification of which is crucial to the understanding of the policy process. The other, as will be seen, is the relationship between ministers and the Prime Minister.
A major reason why departmental civil servants will do their best to assimilate their minister to the existing departmental philosophy and practice, and an important part of their thinking when dragging their heels [Williams 198?], is their position in relation to the ongoing nexus of interests in the policy area with which they are concerned. They are anxious not to disturb current arrangements, and, if forced to do so, are anxious to achieve their end with the minimum of disturbance to good working relationships, even to the point of subtly modulating policy in the course of its implementation. Of course, this pattern of relationships also varies, ranging from those where the department can be said to have emasculated the pressure groups with which it deals [Nettl] to the opposite end of the spectrum where the groups have in effect colonised the department.[?Wilson] Another set of variables affecting departmental policy‑making, and to some extent therefore, Governmental strategy, derive from the nature of the policy area concerned, whether it is highly technical, whether or not the policies operated are highly integrated or diffuse, the extent to which it is affected by exogenous factors, and the extent to which it is characterised by the use of external agencies to implement or apply to a geographical area policies devised at the centre. In health, for example, the way in which provision has been organised and the existence of highly qualified staff are major constraints on the way in which policies are devised and executed in a way which would be much less true of social security, say, yet a Secretary of State for Health who moved on to Education would find himself with far less direct control over policy outcomes. Regulatory agencies, District Health Authorities and local authorities each afford Government different opportunities and problems when compared to direct departmental control and the arrival of "Next Steps" agencies will further complicate the policy making scene.
No Prime Minister, even with enhanced assistance, can hope to control this enormous and varied field, and it is surprisingly easy for the centre of government to lose what little purchase it has over certain areas of policy. Prime Ministers therefore need strong ministers who are respected by their departments and have the energy and ability to drive policies through. Where a John Moore breaks under the strain, a Kenneth Clarke is appointed to take his place, and the administrative scope of the department reduced so that he can concentrate on the task of reform. If the minister concerned can go on to develop authority within the policy community as a whole his job will be eased. Lightweight ministers with purely "ambassadorial skills" are simply not up to the job.
"Mixing the initial cabinet cocktail" and reshaping it periodically to serve the Government's strategic aims, decisions identified by Baldwin as the Prime Minister's main job, are clearly operations which can serve his purposes. It was no accident that, with the exception of the Ministry of Labour, the key economic departments in 1979 went to ministers who were on Mrs Thatcher's wavelength. Because of the various considerations which Prime Ministers must bear in mind, among others a Cabinet broadly representative of the party, upward mobility for the ambitious lest they become restless, administrative ability and propagandist skills, use of placement power rather than simple patronage is the major way for a Prime Minister to further his own policy aims, but he can derive no benefit from such appointments unless the ministers concerned are on top of their job.
The operation of a ministerial hierarchy further constrains choice and, where the ideological preferences of successive party leaders differ, may be of critical importance. It was inevitable that Mrs Thatcher should start with a Cabinet that was not her own. Only two of its members, Biffen and Edwards, had front bench careers entirely of her making: the bringing back of Angus Maude from long internal party exile for a brief spell in the Cabinet was also her doing. By contrast no less than ten of the twenty‑two had first been appointed to Cabinet by her predecessors and two more had been first appointed to the Shadow Cabinet by Heath. That is not to argue that all of them were hostile to her, simply to suggest the difficulty in altering the way in which the career ladder operates within party government. With the exception of Atkins, appointed Chief Whip by Heath in 1974 and now rewarded with Cabinet Office, the remaining Cabinet members had all achieved ministerial rank under Heath and now received their Cabinet appointments from his successor. It was to be twenty months before any changes were made and a new man, entirely of Mrs Thatcher's making, joined the Cabinet in the person of Leon Brittan. Eight months later in September 1981 she felt able to replace three members of the Cabinet with people entirely to her liking, but even with three further vacancies caused by resignations (not dismissals) only eight of her twenty‑two strong Cabinet at the time of the 1983 election owed their entire front bench career to her. By then she had been leader of her party for more than eight years and Prime Minister for four.
For any Prime Minister therefore, there are trade‑offs between picking men whose views are congenial to him, the necessities of party management, and the choice of those who can make their views stick in the policy arenas to which they are despatched. If, in general, Prime Ministers prefer competence to ideology, that merely reflects their desire to be re‑elected.
Both resignations and sackings are politically damaging in almost all circumstances and therefore to be avoided where possible, but it is not often realised that from the Prime Ministerial angle frequent changes of minister are scarcely less damaging. Ministers cannot be expected instantly to impose their will on their new department. Civil servants will have a chance to re‑open arguments previously lost and the minister, however experienced, needs time to master his new responsibilities. Once the change has been made, the new man must be given at least eighteen months to make his mark. While placement is a genuine Prime Ministerial power, therefore, it can be used in relation to each department at best twice in the course of a Parliament and there will be a price to be paid.
The use of Cabinet Committees can be helpful to a Prime Minister, more particularly if he takes the chair, but rather less so than Crossman supposed [Crossman 1972]. It may be possible for an ad hoc committee to be stacked, particularly where there are not a great many departmental interests to be accommodated, but the standing committees of the Cabinet have to incorporate all the relevant departments if their decisions are to command acceptance. Indeed it can be argued that the Prime Minister is more powerful in full Cabinet than he can ever be in committee. Perhaps half‑a‑dozen among the party's leaders command a place in cabinet. The remainder depend on Prime Ministerial patronage, not so much for their place in government, but for the particular rank they hold. Given that they are to this extent substitutable, they may well defer to the Prime Minister in Cabinet.
In committee, where most ministers have a departmental brief, the luxury of such support is less available. As Mrs Thatcher clearly realised at the outset of her administration, the Prime Minister will gain more influence by making the appropriate appointments to particular posts than ever she will from adding members of her choice to Cabinet committees, but almost inevitably there will be few such men on whom the Prime Minister can wholly rely and who will nonetheless have the capacity to effect his bidding.
However, if Prime Ministers need strong, competent ministers, they in turn are conscious of the need for a strong lead from the centre. In fact, it could be argued that the best cards the Prime Minister has in his hand, apart from an ability to deliver success at the polls, are the general desire within Whitehall for coherence in Government policy and a sense among ministers that there has to be some overall strategy, not least because the winning of elections is seen to be a medium‑term activity. Neither are possible without a considerable degree of integration in the business of government, and while the Treasury can and does offer its own brand of coherence, its ability to implement anything more than an economic strategy is impaired by the fact that it has departmental axes of its own to grind. In the past the Cabinet itself, but increasingly more in posse than esse, was a major factor in achieving integration. As early as the 1920s, it had already come to be the conference of busy departmental heads so brilliantly described by Leo Amery [Amery 1935], and it should come as no surprise that most of her ministers welcomed Mrs Thatcher's decision to hold only one Cabinet meeting a week. Perhaps because of the opposition she encountered in a cabinet not of her making [Prior 198x], Mrs Thatcher ensured that more and more of its business was transacted in Cabinet Committee and in still more smaller groups, not all of which were even dignified by the attentions of the Cabinet Secretariat. There were resurgences of Cabinet discussion whenever the Prime Minister was in difficulties over her leadership, eg after Westland, and certain key matters like the ratification of the final outcome of the public expenditure process and the initial settlement of the total within which bargaining took place have always been taken in full Cabinet. But towards the end of Mrs Thatcher's term few issues were being taken to Cabinet with accompanying papers and the meetings increasingly resembled a weekly stocktaking on the Parliamentary and political scene. Conceivably it was no less valuable for that.
Ministers have to attend cabinet and yet in the past they have had little or no direct concern with much of its business. Cabinet Committees have come to seem much more relevant to their own concerns, and even there, if the matters on the agenda look stultifyingly dull or unimportant, they can send a junior minister to represent them.
Although it would be wrong to suggest that, even before the war, nothing was settled in this way, more now is simply agreed by ministerial correspondence. The integrating role, once played by the Cabinet, is today played by the Cabinet Office, Cabinet Committees and the Prime Minister in that order of importance. Although the Prime Minister's attention, at least on any continuous basis, can only be given to a relatively small number of areas, they are the most key in strategic terms.
Ministers are conscious of the need for a grand strategy, which is shaped as much, if not more, by political considerations as the needs of government and that this requires someone to take a synoptic view of what they are about. They know that only No 10 can play this role and are ready therefore to defer a great deal to the Prime minister's thinking, more particularly when this appears to be eliciting the right response from the nation. Practical rather than ideological considerations are uppermost in their mind, and there is an imperative to seek some form of accommodation. What is instructive about the resignations of Lawson and Howe is the length of time over which their evident differences of opinion with the Prime Minister were composed, at first by their accommodation to her view, but latterly as the political going got rough, by her accommodation to their views. Her attempt to ease this situation by shifting Howe from the Foreign Office, seen initially as a successful exercise in power, put too great a strain on their relationship, but until that moment, nothing could better have illustrated both the symbiotic relationships at the centre of government and the enormous premium placed on a strategy which will hold the party together and carry it towards an election victory. The Prime Minister is therefore more than the "guardian" of the strategy: s/he is its chief architect.
Cautionary words are in order about the extent to which this command over broad strategy can be translated into detail. Quite apart from time constraints, Prime Ministers have to carry the relevant departmental ministers with them, and that, it should be observed, is particularly true when they are using bilaterals or informal groups of ministers to speed business. The decision to take the matter forward in these circumstances is the departmental minister's.
Given these complex and shifting relationships within the policy making model, the need to specify it more precisely and to identify the relative degree of stability in the patterns and the reasons for variation is a major research task. However preliminary examination suggests that all can be subsumed under a number of inductively derived "If.... then" propositions, even if their status has to be that of a law‑like generalisation rather than a general law. The constitutional rules within which the political game is played at the heart of the machine are clearly a major stabilising factor, and their existence makes it not only plausible but almost certainly right to continue to categorise the model as one of "Cabinet government". The focus of all political activity is the attempt to alter both resource patterns and the pattern of power, but the instrument used to legitimate such attempts, Parliament, is under the control of the leaders of the majority party as long as they operate with the consent of their backbenchers. Despite all pressures to entropy, no other part of the policy making machine nor the groups which attempt to suborn it can secure the necessary changes without recourse to Parliament, and in practice that must mean the Government of the day.
What ties the policy making machine together and gives it what coherence it has in the face of a large number of conflicting pressures is a single key institution, the Cabinet. While this institution is more properly characterised as a complex mechanism rather than a weekly meeting, its importance cannot be questioned for two main reasons.
First and foremost, it transforms political authority derived from leadership of the party into constitutional authority within the government machine. To illustrate the point: one of the key co‑ordinating mechanisms, the Cabinet Office works to the Prime Minister because he chairs the Cabinet. If the Cabinet secretary can be plausibly described as the Prime Minister's Permanent Secretary, as Wilson suggested [], it is not because the Prime Minister is leader of his party, but because that party has a majority in Parliament and the leader as a result is chairman of the Cabinet. The bulk of the executive authority conferred on government is formally vested in neither Cabinet nor Prime Minister but in the various departmental ministers. However the convention of collective responsibility operates to give the Cabinet predominance. When a minister argued otherwise on the basis of the customary form of its minutes, Attlee tersely ruled: "The Cabinet does not advise. It decides." [Harris 197x] The departments themselves acknowledge that a Cabinet minute is binding on them. Cabinet committees and their sub committees derive their authority from the Cabinet and possess it only because the Cabinet exists. To that extent the point made by Mackintosh that many decisions are taken elsewhere loses all force. With the exception of decisions made on the authority of the departmental minister alone, decisions can be taken elsewhere and have binding force only because of the existence of a central body which legitimises all forms of collective decision taking and to which the individuals concerned can appeal if they dislike the decision reached. This is clearly exemplified in the public expenditure process. The bilaterals which take place and the effective veto wielded by the Chief Secretary in Cabinet committee [Barnett 1982] depend in the last resort on the Cabinet's authority, and while the final ratification of the Autumn statement by the Cabinet is usually a formality, discussion in July of the planning total which governs the whole process most certainly is not.
While most commentators ‑ even those seeking to minimise the importance of the institution ‑ note the right of appeal by ministers to Cabinet, all seem to have failed to grasp the central importance of the Cabinet to the Prime Minister. It is something more even than the "steering wheel by which he steers the Governement" [Armstrong 1975], although that observation contains an important part of the truth. The Westland case illustrates the point. Mrs Thatcher's preference for small, informal groups is well documented, whether they were used, as Lord Armstrong suggests, to clear her own mind prior to the meeting of other more formal bodies [Armstrong 1990] or for the purpose of decision‑taking as others contend. In a sense the point is immaterial. If such meetings were to result in a decision, it had to be that of the relevant departmental minister since such meetings have no formal status. The Prime Minister could secure her way at such meetings only by securing the assent of the minister who had to implement the decision. In the case of Westland, Heseltine, when asked to withdraw the National Armaments Directors' agreement reached earlier at his behest, insisted on taking the matter to a Cabinet committee, EA on 9 December 1985. The minuted decision of that committee was held by the Prime Minister and Cabinet Secretary, if the Westland Board rejected the "European offer", to have obviated any need for a further meeting on the subject.
Heseltine disagreed, and when he learnt that there would be no second meeting, he attempted to raise the matter orally in Cabinet on 12 December, and subsequently asked for his protest to be recorded in the minutes. Whether or not Heseltine was correct is not germane to this argument. The matter was discussed by the full Cabinet on 19 December and a decision reached along the lines indicated by the Prime Minister, that it was for the Westland Board to decide what was best for the company and that ministers should avoid further public comment. Although Heseltine reserved his right to answer factual questions, he acquiesced in the decision. However when the House was told, he showed visible dissent. There followed a torrent of press leaks and briefings by the two departments concerned, a strong indication from the Defence Ministry that if Westland accepted the American offer it might be excluded from European collaboration, and, in response to an alarmed request from Westland for clarification of the Government's position, a carefully crafted letter which embodied both Heseltine's indication of difficulties on the European front and reassurances from the Prime Minister that the Government would stand by the company whatever decision it reached. The details of the further exchanges, including the notorious leak of the Solicitor General's letter, which contained criticisms of a further letter written to the European consortium by Heseltine, are of interest in this context only because they suggest the extent to which ministers had gone in internecine warfare.
Once again the Prime Minister had to resort to the Cabinet to enforce collective responsibility. When the Cabinet agreed on 9 January that all statements and answers on Westland must in future be cleared through the Cabinet Office, Heseltine resigned [Linklater and Leigh 1986].
The rights and wrongs of Westland need not be further debated here. What this example underlines is that when a Prime Minister wished to enforce her view on a dissident minister, short of sacking him outright, the only way in which she could do so was by seeking an authoritative ruling from the Cabinet, something which she did on 19 December 1985 and 9 January 1986. Similarly Eden had recourse to his Cabinet on 28 August 1956, when there were obvious signs of dispute amongst his ministers, in order to secure an agreed line to which they could all subscribe [Barnes 1990].
Although it neither specifies the precise balance of power between the Prime Minister and his departmental ministers nor between ministers and officials at any given point of time, characterising British government as Cabinet Government encapsulates the most salient point about its conduct. The dynamics of the system are governed by a framework of rules derived from constitutional principle. Protean though the system is, the precise balance between its elements is constrained by those rules and will ultimately be registered, if not struck, in Cabinet. The term remains the best way of characterising the system as a whole, therefore, but argument on this limited point should now give place to the need to explore in detail the elements of a dynamic model of Cabinet Government and the precise circumstances which give rise to varying distributions of power within it.