Policy Papers

The power of the Prime Minister

Andrew Blick , George Jones |

  • RSS Feed Icon

Executive Summary

  • Two recent reports by influential bodies have advanced the theory that the British premiership has recently become more dominant, even developing into a presidency.
  • However, such claims have been made repeatedly over the past three centuries and a historical perspective shows that claims about the greater dominance and presidentialism of the premiership should be treated with caution.
  • No.10 has always been subject to intense media attention, often focusing on the person of the Prime Minister, and has achieved only limited success in managing the way it is portrayed.
  • Under Tony Blair, a semi-official 'Department of the Prime Minister' was established, but this change can be seen as a reversion to the pre-mid-nineteenth century arrangements when premiers usually had their own 'department' in the form of the Treasury.
  • The possession of a department should not be equated directly with greater dominance for the Prime Minister.
  • The recent formation of a Conservative-Liberal Democrat coalition government demonstrates that, rather than progressing steadily towards greater dominance, No.10 operates within a dynamic framework for the exercise of power that can change rapidly and unpredictably.
  • History suggests that the decision-making processes of collective cabinet government may prove particularly valuable in the circumstances of a two-party coalition.


The role of the British premiership has, since the office began in the early eighteenth century, often been a prominent subject of political debate. This year two reports, by the House of Lords Constitution Committee and the Institute for Government, have considered this topic. Both are likely to have some influence upon perceptions of the office of Prime Minister and the way those within and outside government approach it.

The following article draws upon our ongoing research into the premiership and aides to the Prime Minister to assess these publications from a historical perspective. We do so partly on the grounds that any discussion can potentially be enhanced through this technique, and because the works under examination advance specific theories about developments over time, which by their nature invite consideration by historians. Having conducted this assessment, we set out some general recommendations for those within government involved in operating and dealing with the British premiership.

Prime ministers, presidents and repeated claims

Central to the interpretations of the premiership in both reports are two related ideas. First, in The Cabinet Office and the Centre of Government, the House of Lords Constitution Committee contends that 'Our evidence suggested that the role of the Prime Minister has changed'. It quotes a series of statements from witnesses about the supposed development of a 'more dominant Prime Minister'. Second, the Institute for Government report, Shaping Up: A Whitehall for the Future, contains a section headed 'The problem of presidentialism', and endorses the notion that the office of Prime Minister has been morphing into a presidency.

Using an historical perspective, it is possible to compare these theses of the premiership with those advanced in earlier periods. Claims about a hegemonic No.10 similar to that set out by the Constitution Committee are both as old as the premiership itself and prone to recur. Robert Walpole is commonly regarded as having become the first Prime Minister during his period of political pre-eminence during the early eighteenth century. His ally, John, Lord Hervey, later wrote of Walpole that 'he did everything alone...whilst those ciphers of the Cabinet signed everything he dictated...without the least share of honour or power'. In 1741 Samuel Sandys criticised Walpole in the Commons for having supposedly achieved 'the sole direction of all public affairs'.

Similar comments were made about subsequent prime ministers. Lord Grenville, who succeeded William Pitt the Younger as premier after his death in 1806, described Pitt as having led in his second, final period of office 'a Cabinet of cyphers and a government of one man alone'. During his premiership of 1828-30 the Duke of Wellington was described by an ally as 'sole Minister and decidedly superior to all'. One critic said Wellington's ministers 'dare not have an opinion, but must move either to the right or the left as this Dictator may think proper'. The diarist Charles Greville referred to Wellington's 'ministerial despotism'.

In the early twentieth century, Sidney Low argued in The Governance of England that the 'Prime Minister's influence and importance are growing' and 'Much of the authority of the Cabinet has insensibly passed over to that of the Premier'. Half a century later Harold Laski, talking of Attlee at the head of the first majority Labour government after the Second World War argued that 'if we compare 1850 with 1950, or even 1900 with 1950, the centralisation of power in the Prime Minister's hands has proceeded at a swift pace'. In 1963 Richard Crossman, then Labour's education spokesman, claimed: 'The post-war epoch has seen the final transformation of cabinet government into prime ministerial government', while in 1968 John Mackintosh wrote that 'the politics of the 1960s have strengthened rather than weakened or altered the lines of development which have led contemporary British Government to be described as Prime Ministerial rather than Cabinet Government'.

According to Tony Benn, in a lecture given in 1979:


the wide range of powers at present exercised by a British Prime Minister...are now so great as to encroach upon the legitimate rights of the electorate, undermine the essential role of Parliament, usurp some of the functions of collective Cabinet decision-making, and neutralise much of the influence deriving from the internal democracy of the Party.


The repetition of such similar claims over such a long period of time means we should treat with circumspection the Constitution Committee's reference to a 'more dominant Prime Minister' today. In earlier eras it was common to regard No.10 as supreme within government, and possibly as becoming ever more so. If such interpretations have at any point in the past been correct, then the idea that the premiership is becoming 'more dominant' is harder to sustain. Could it be possible, for example, to attain greater 'dominance' than that associated with Pitt the Younger's 'government of one man alone', or the alleged 'Dictator' Wellington? On the other hand, if the present-day claim put forward by the Constitution Committee is accurate, then all the earlier theories described above are, by implication, dubious.

For similar reasons the 'presidentialism' thesis advanced in the Institute of Government report must be approached with scepticism. The Institute argues that 'the presidentalism debate is at least 40 years old'. It is correct, but understates the case.

Low argued in The Governance of England (1904) that for 'the greater part of the past half century...The office of Premier has become more than ever like that of an elective President'. An article published in October 1920 by the eminent political scientist, Harold Laski, argued that David Lloyd George, rather than being chair of the Cabinet, had become 'virtually the President of a State'. In 1963 Crossman wrote that by the mid-twentieth century prime ministers wielded 'near-presidential powers'. George Brown resigned as Foreign Secretary in 1968 because, he believed, the premier Harold Wilson was 'introducing a "presidential" system into the running of the Government'. If any of these earlier claims are correct then it is difficult to sustain the argument that a UK 'presidency' has developed recently.

In the following sections we consider the evidence for the supposed increase in prime-ministerial dominance.

The international role of prime ministers

To support its contention that the Prime Minister is now more dominant, the Constitution Committee quotes the Cabinet Secretary and Head of the Home Civil Service, Gus O'Donnell stating that: 'The number of overseas visits for the Prime Minister has gone up. That is a trend of globalisation. Prime Ministers inevitably are going to be much more involved in that global role and I think that is important.'

The number of trips abroad made by premiers may well have risen. But this trend in itself does not establish that they are 'much more involved in' a 'global role'. Before the invention of the aeroplane, prime ministers such as William Pitt the Elder (the Earl of Chatham) in the mid-eighteenth century, Lord Palmerston in the mid-nineteenth century and Lord Salisbury at the turn of the nineteenth and twentieth centuries (who combined the post of premier with that of Foreign Secretary) all played a highly pronounced role in foreign affairs, which can be said to be more truly 'global' as in all these eras the UK was a first-rank world power. Benjamin Disraeli's participation in the Congress of Berlin of 1878 was more momentous than, for example, Gordon Brown or David Cameron flying to Afghanistan, largely for publicity purposes, to spend a few hours meeting British troops.

No.10 and the media

Lord Lipsey, formerly a special adviser to Prime Minister James Callaghan in the late 1970s, is also quoted by the Constitution Committee:


the media did not, in our day, hold the Prime Minister responsible for every single thing that happened in every single corner of Whitehall ... and there was not need for the Prime Minister to react swiftly to everything that happened, as present Prime Ministers have to. I think that is a very strong pressure which tends in the direction of a more prime ministerial system.


The Institute for Government report makes a similar claim, arguing that that 'the political personality of the PM is increasingly important' and contrasting the relative levels of media interest in Clement Attlee and Margaret Thatcher.

While there have been changes in the way the media operate - associated in particular with technological developments - No.10 has always been subject to intense public scrutiny, often focusing on the person of the Prime Minister, and has felt obliged to respond to it, though only ever with limited success.

The pattern was set under Walpole. Few other politicians have endured such concerted attacks of the quality launched against him by the 'Scriblerus Club' - a collection of Tory authors of literary talent unsurpassed in any era, excluded from office and favour by Whig hegemony. Parodies and criticisms of Walpole are contained in works including Gulliver's Travels by Jonathan Swift, Alexander Pope's Dunciad and John Gay's The Beggar's Opera. They all treat Walpole the person as synonymous with his government, an assault on the former serving as an attack on the latter. In The Beggar's Opera the famous couplet 'How happy could I be with either, Were t'other dear charmer away' was understood as a depiction of the triangular relationship between Walpole, Lady Walpole and Walpole's mistress, Maria Skerret. Swift made similar insinuations in Gulliver's Travels, and in a poetic tirade of 1738 Pope stated: 'Sir ROBERT'S mighty dull, Has never made a friend in private life, And was, besides, a tyrant to his wife'.

Walpole responded to this onslaught by constructing a political communications operation to rival that of any subsequent Prime Minister. At first the methods were defensive. Control of the postal service was used to prevent the distribution of critical works and writers were bribed not to author them. There then began a campaign on a grand scale, involving the production, printing and circulation of newspapers, journals and poems praising Walpole and attacking his enemies.

These tendencies continued throughout the centuries that followed and it is hard precisely to establish whether they were more intense in one era than another. There may be a difference in the levels of media coverage of Attlee and Thatcher. But rather than proving an overall historical trend, this discrepancy could be seen as a consequence of the differing style of their premierships, with the latter more domineering and attention-grabbing than the former. Such variability can be detected throughout the development of the office of Prime Minister, with frequent alternations between more and less forceful premiers: such as Walpole and the Earl of Wilmington, the Earl of Aberdeen and Palmerston, Herbert Asquith and David Lloyd George, John Major and Tony Blair.

A 'Department of the Prime Minister'?

The Constitution Committee notes that Lord Butler, Cabinet Secretary and Head of the Home Civil Service from 1988-98, described how in 1997 'it was part of the explicit purpose of Mr Blair to strengthen the centre, and... to make the Cabinet Office a part of the Prime Minister's Department'.

In our previous History & Policy paper, The 'Department of the Prime Minister' - should it continue? we discussed the view that Blair established a Department of the Prime Minister into which the Cabinet Office was incorporated. Lord Butler's argument has substance, since the terms of reference of the Cabinet Office were progressively changed to eliminate by 2002 any reference to the provision of assistance for collective Cabinet government, and a less traditional role of supporting the Prime Minister as leader of the government was emphasised. The number of staff serving the Prime Minister grew to become closer to a total that might be expected of a full department of state. It reached a peak in 2005-6 of 782, which exceeded the figure from any previous era by more than 500, that is over 300 per cent. It dropped later but in 2007-8 still stood at 399, higher than at any time before 1997. In addition the staff attached to No.10 took on more active and interventionist roles in the development and implementation of a wider range of policies, from international to local affairs, and in the organisation of the bureaucratic machine.

The Institute for Government on the other hand insists that 'the Cabinet Office is a long way from becoming a fully fledged premier's department'. This claim is correct in the sense that the Prime Minister largely lacks the direct policy responsibilities, either in statute or by convention under the Royal Prerogative, possessed by secretaries of state, who have substantial budgets voted to them by Parliament. But greater progress has been made towards the establishment of a 'department' than at any point in the last 150 years.

While the establishment of a semi-official Department of the Prime Minister is a change, it is to some extent of a cyclical nature. In an 1803 description of the post of Prime Minister - or as he put it the 'avowed and real Minister' - Pitt the Younger noted, 'that Minister ought...to be the person at the head of the finances'. During the early period of the premiership, in the later eighteenth and early nineteenth centuries, the individual recognised as Prime Minister was almost always designated 'First Lord of the Treasury', a role which until the mid-nineteenth century brought with it direct responsibility for the Treasury. If they sat in the Commons, First Lords were even more closely involved with the Treasury and its business since they held this post jointly with that of Chancellor of the Exchequer (this post was not yet considered a first-rank position in its own right) and delivered the Budget speech. The premier was in this sense a departmental minister with a specific policy portfolio and relatively sizeable staff, as in the nascent days of the premiership the Treasury was by far the largest office of government.

In 1841 Robert Peel took up the post of First Lord of the Treasury, but not that of Chancellor of the Exchequer. Peel planned to delegate the detailed work of the Treasury to the Chancellor of the Exchequer, enabling him as Prime Minister to range across the whole of government. Peel's premiership retained ultimate control over financial policy and engaged closely in it when he deemed it necessary. The long-term consequences of his innovation were that the Chancellors of the Exchequer emerged as an office more clearly distinct from that of the First Lord, and the leadership and management of the Treasury was ceded by the First Lord to the increasingly prominent Chancellor.

The historian Maurice Wright observed that, up to the mid-nineteenth century, 'the First Lord [was] intimately connected with Treasury business', but 'after 1856...normally took no part in the formal transaction of Treasury business, unless the office was combined with that of Chancellor of the Exchequer'. In 1893 Algernon West, the senior private secretary to William Gladstone, described the position as it had developed in the post-Peel era: 'Departmentally the First Minister of the Crown and the head of the Government has nothing to do'. Premiers were, 'as the French say, "without portfolio"', since 'the First Lord of the Treasury has nothing to do with the Treasury. His presidency is merely nominal, the real headship being in the hands of the Chancellor of the Exchequer'.

The old nature of the premiership was remembered for a while. When Gladstone temporarily recombined the office of Prime Minister with Chancellor in August 1873, The Times noted his decision to 'be his own Chancellor of the Exchequer' was 'a plan equally simple and unexpected, although not novel...This was common enough in past generations, and we are returning in many ways to the practice of the past'. But it was eventually forgotten that the premier was once a 'departmental' minister.

The office of Prime Minister had left its embryonic first phase as a 'finance ministry-plus' and moved into a second phase, in which the role of overall government coordination was of heightened importance. Between the mid-nineteenth century and the Blair premiership, the non-departmental nature of the premiership broadly persisted, but, as Lord Butler's comments suggest, from 1997 there were signs that a third phase of the office was emerging, with the large-scale staff associated with phase one used to augment the cross-departmental role that had become more prominent in phase two.

Does a swollen support staff combined with more interventionist roles necessarily bring about a 'more dominant Prime Minister'? The experience of early prime ministers such as Pitt the Younger was that responsibility for the Treasury could be a hindrance as well as a help. Four decades later Robert Peel was by any standard an interventionist Prime Minister, who seemingly chose to distance himself from the management of a 'department' in order to dominate government more, not less.

One reason the second, non-departmental, phase of the premiership has persisted is that some Prime Ministers feared that formally establishing a 'Department of the Prime Minister', rather than simply strengthening No.10, might create various problems. Harold Wilson initially contemplated such an innovation but ultimately decided against it. In explaining his decision Wilson made reference to arrangements in the US:


The president...can hardly move for staff. (In President Lyndon Johnson's time, I was told his staff was 2500: I understand it increased still further under President Nixon.) He is pressed on all sides for signatures, approvals, ratifications - I have seen presidents badgered to sign them in the lift, an action that must be a more or less automatic reflex.


Simply possessing a large staff with interventionist functions cannot necessarily overcome the prevailing political circumstances, such as those faced by David Cameron today, who must accommodate a second party within government. We do not suggest that the Cameron premiership will necessarily be weak, but that it cannot achieve its goals simply by endowing itself with an enlarged support structure.

Measuring Cabinet government

The Institute for Government claims in support of the 'presidentialism' thesis that 'there is significant evidence that the role of Cabinet as a formal decision-making body has been in decline since the war', citing data showing a decline in the number of full Cabinet meetings.

While these statistics are significant, caution should be exercised before equating them with some form of 'presidency'. Analysis of the table below demonstrates the difficulties. It compares data from 1921, the last full year of the Lloyd George premiership, and from 1923, when his successors, Andrew Bonar Law and Stanley Baldwin, occupied No.10.


Year/premier 1921/Lloyd George 1923/Bonar Law/Baldwin
Full Cabinet meetings 93 59
Cabinet sub-committee meetings 220 110
Cabinet memoranda 1,200 424
Combined staff employed by Cabinet Office and Committee of Imperial Defence 143 39

The drops recorded in all these figures between 1921 and 1923 could be read as suggesting the Lloyd George premiership was far more collegiate than those of Bonar Law and Baldwin, with Lloyd George being more committed to holding meetings, circulating papers and furnishing staff to support collective processes. Yet first-hand accounts of government from the time suggest that in practice, Lloyd George's No.10 - likened by Laski to a presidency - was more domineering than that of his successors. Measurable indicators are important but cannot tell the whole story. The mere occurrence of a meeting, circulation of a paper or existence of a secretariat does not reveal the quality or dynamics of the discussions that took place. It may be that the full use of collective procedures was a means by which Lloyd George reconciled his dynamic style with the operation of a Conservative-Liberal coalition government - a lesson for David Cameron.


Claims about an increasingly dominant or more presidential premiership have a very long lineage, but there are difficulties in sustaining them. Why have they been made so often?

One possible reason is ignorance about debates of earlier times that only historical research can correct.

Secondly, such assertions may have been deployed repeatedly over time as a means of political attack, by anyone seeking to criticise the style or personality of a particular Prime Minister.

Thirdly, such assessments of the premiership may be attractive because they provide a dramatic - if simplistic - narrative. Images of unprecedented or accelerating developments occurring in the present and pointing only in one direction can be easier and more compelling to convey than more nuanced accounts.

Fourthly, it may be a means by which those, within and beyond the Cabinet, can distance themselves from particular government actions - such as the Iraq war - with which they do not wish to be associated, claiming they could not possibly resist.

Fifthly, there has always been the potential for the premiership to have a substantial impact upon government, and tensions can often develop between the public-leadership role of the Prime Minister and the practice of collective Cabinet government resting on parliamentary consent. The former is sometimes mistaken for growing dominance or presidentialism.

Prime-ministerial predominance fluctuates with changing circumstances. Taking the long sweep from 1721 to 2010, we have shown that its manifestations are not permanent political realities, but temporary and contingent. Events, such as the formation of a coalition government in May 2010 could well mean that the Prime Minister has no choice but to operate in a more collegiate fashion, as is set out explicitly in the recently-published 'Coalition Agreement for Stability and Reform'. David Cameron will find it useful to bind the whole Cabinet, Conservative and Liberal Democrat members alike, into unpopular decisions about public-spending cuts, thereby sharing the blame. This may explain the significant roles - and tough responsibilities - awarded to Liberal Democrats, such as Vince Cable as Business Secretary and Danny Alexander as Chief Secretary at the Treasury.

A number of recommendations follow from our analysis:

  • Modern Prime Ministers should not blindly accept claims that the office is becoming ever more dominant or presidential. The way they conduct themselves should take into account the contingent nature of current 'dominance' and the limitations that exist. The more enduring doctrines of collective cabinet and individual ministerial responsibility remain essential components of the British Constitution. British government cannot be run from one office, particularly in the current circumstances of a coalition government.
  • No.10 must accept the inevitability of hostile and personalised media criticism directed at the Prime Minister, which can be managed only to a limited extent.
  • When assessing the nature and size of staff-support needed, whoever occupies No.10 should take into account the political, administrative and logistical problems that can come with a body akin to a full-sized government department. By scaling-down, a Prime Minister may achieve more.
  • The operation of Cabinet government cannot be understood purely by the numbers of meetings held and papers circulated. Any attempt to ensure the effective functioning of Cabinet should be based on an understanding of collective government, taking in qualitative as well as quantitative considerations. Ministers who sit with the Prime Minister in Cabinet should recognise that - under existing constitutional understandings - they are able and required to assert their rights as members of a collective rather than simply receive instructions from No.10. The ultimate sanction available to them is resignation. They cannot use the excuse of a supposedly dominant or presidential premiership as a means of distancing themselves from decisions with which they disagree.

Further Reading

Peter Hennessy, The Prime Minister: The office and its holders since 1945 (London: Allen Lane, 2000)

House of Lords Constitution Committee, The Cabinet Office and the Centre of Government, 4th Report of Session 2009-10, HL 30 (London: The Stationery Office, 2010)

Institute for Government, Shaping Up: A Whitehall for the future (London: 2010)

Paul Langford, 'Prime Ministers and Parliaments: the long view, Walpole to Blair', The Annual History of Parliament Lecture, 2005, Parliamentary History, 25 (2006), pp. 382-94

Tone Sundt Urstad, Sir Robert Walpole's Poets: The use of literature as pro-government propaganda, 1721-1742 (London: Associated University Presses, 1999)

Maurice Wright, Treasury Control of the Civil Service: 1854-1874 (Oxford: Oxford University Press, 1969)


Papers By Author

Papers by Theme


Sign up to receive announcements on events, the latest research and more!

To complete the subscription process, please click the link in the email we just sent you.

We will never send spam and you can unsubscribe any time.

About Us

H&P is based at the Institute of Historical Research, Senate House, University of London.

We are the only project in the UK providing access to an international network of more than 500 historians with a broad range of expertise. H&P offers a range of resources for historians, policy makers and journalists.

Read More